Article(id=1297211746784141990, tenantId=1146029695717560320, journalId=1296125453100220459, issueId=1297211624738284246, articleNumber=null, orderNo=null, doi=10.11975/j.issn.1002-6819.202601127, pmid=null, cstr=null, oa=null, hot=null, price=null, onlineType=0, articleFormat=0, articleType=null, articleTypeStr=null, receivedDate=1768492800000, receivedDateStr=2026-01-16, revisedDate=1775318400000, revisedDateStr=2026-04-05, acceptedDate=null, acceptedDateStr=null, onlineDate=1787208981461, onlineDateStr=2026-08-20, pubDate=1782748800000, pubDateStr=2026-06-30, doiRegisterDate=null, doiRegisterDateStr=null, onlineIssueDate=1787208981461, onlineIssueDateStr=2026-08-20, onlineJustAcceptDate=null, onlineJustAcceptDateStr=null, onlineFirstDate=null, onlineFirstDateStr=null, sourceXml=null, magXml=null, createTime=1787208981461, creator=13701087609, updateTime=1787208981461, updator=13701087609, issue=Issue{id=1297211624738284246, tenantId=1146029695717560320, journalId=1296125453100220459, year='2026', volume='42', issue='12', pageStart='1', pageEnd='396', issueExtLink='null', onlineDate='null', pubDate='1782748800000', pubDateStr='2026-06-30', beforeIssueId=null, nextIssueId=null, price=null, status=1, issueComplete=1, articleOrder=1, issueType=1, specialIssue=null, createTime=1787208952364, creator='13701087609', updateTime=1787212261177, updator='13701087609', preIssue=null, nextIssue=null, articleTotal=null, ext={EN=IssueExt(id=1297225503002357852, tenantId=1146029695717560320, journalId=1296125453100220459, issueId=1297211624738284246, language=EN, specialIssueTitle=, coverIllustrator=null, specialIssueEditor=, specialIssueAbout=), CN=IssueExt(id=1297225503002357853, tenantId=1146029695717560320, journalId=1296125453100220459, issueId=1297211624738284246, language=CN, specialIssueTitle=, coverIllustrator=null, specialIssueEditor=, specialIssueAbout=)}, issueFiles=null, downloadFileDto=null}, startPage=384, endPage=396, ext={EN=ArticleExt(id=1297211747060966056, articleId=1297211746784141990, tenantId=1146029695717560320, journalId=1296125453100220459, language=EN, title=Dual-model progressive feeding strategy for Micropterus salmoides, columnId=1297211746972885671, journalTitle=Transactions of the Chinese Society of Agricultural Engineering, columnName=Agricultural Produce Processing Engineering, runingTitle=null, highlight=null, articleAbstract=

Accurately adjusting the feeding intake is often required in pond aquaculture of Micropterus salmoides. In this study, a dual-model progressive feeding strategy was proposed to accurately forecast the postprandial feeding status of the subsequent round using a time series model. The decrement mechanism was first triggered. Subsequently, a classification model was utilized to determine the postprandial feeding status of the current round, enabling the termination of the decrement mechanism. The experimental system consisted of three pond culture tanks stocked with Micropterus salmoides, with initial body masses of (161.47±11.02) , (204.36±17.09) , and (220.25±22.78) g, and the stocking densities of 224, 301, and 294 individuals, respectively. Feeding experiments were conducted using a multi-round feeding protocol within a single session. Audio data was collected during feeding using the digital hydrophone, while video data was obtained using a camera. Meanwhile, multimodal data including light intensity, water temperature, body mass, per-round feeding rate, and population abundance were synchronously recorded for per-round feeding. The stabilization of the cumulative feeding energy curve was adopted to determine feeding termination. A dynamic adaptive threshold segmentation was employed to accurately identify the feeding audio within the collected audio. Principal component analysis was conducted on the feeding features to extract from the feeding audio. Feature selection was employed to identify sensitive features to feeding status, which served as the inputs for model training and classification. Four models 1D convolutional neural network, Long short-term memory, gated recurrent unit, and transformer were improved to construct the feeding prediction models with time series. The optimal model was selected to predict the postprandial feeding status of the subsequent round, serving as the first-stage model of feeding strategy. Eight models k-Nearest Neighbors, decision tree, support vector machine, random forests, adaBoost, gradient boosting decision trees, extreme gradient boosting, and light gradient boosting machine (LightGBM) were employed to construct feeding status classification models. The best classification performance was selected as the feeding determination model in the second stage. Results showed that the sensitive features included overall features, light intensity, water temperature, body mass, and per-round feeding rate. A classification between the original and sensitive features demonstrated the effectiveness of the sensitive features, according to gradient boosting decision trees and random forests models. Transformer-regression-classification (transformer-RC) model outperformed the rest of the models across 1 to 3-time windows, with the accuracy from 0.93 to 0.94. The Transformer-RC model effectively predicted the postprandial feeding status of the subsequent round. All five ensemble models achieved strong classification, which outperformed the three base models. Among them, the LightGBM model achieved an accuracy of 0.98 for inputs to the classification model of feeding status in the second stage of the feeding strategy. Both the Transformer-RC and LightGBM models shared high prediction and classification after validation, with average values of four evaluation metrics exceeding 0.89 under both feeding states, indicating strong generalization. This finding can provide a strong reference to develop intelligent feeding.

, authors=Yaping LI, Hequn TAN*, Yifan CHEN, Yiren ZHANG, authorsList=Yaping LI, Hequn TAN, Yifan CHEN, Yiren ZHANG, authorCompany=null, correspAuthors=Hequn TAN, authorNote=null, correspAuthorsNote=null, copyrightStatement=Copyright © 2026 Transactions of the Chinese Society of Agricultural Engineering., copyrightOwner=null, extLink=null, articleAbsUrl=null, sourceXml=null, magXml=null, pdfUrl=null, pdf=null, pdfFileSize=null, pdfExtLink=null, richHtmlUrl=null, mobilePdfUrl=null, reviewReport=null, pdfFirstPage=null, abstractGraph=null, abstractGraphContent=null, abstractVideo=null, citation=null, cebUrl=null, magXmlContent=null, mapNumber=null, fund=null), CN=ArticleExt(id=1297211750953280194, articleId=1297211746784141990, tenantId=1146029695717560320, journalId=1296125453100220459, language=CN, title=基于双模型渐进式的大口黑鲈投喂策略, columnId=1297211747186795177, journalTitle=农业工程学报, columnName=农产品加工工程, runingTitle=null, highlight=null, articleAbstract=

针对大口黑鲈养殖过程中难以精确且自动调整投喂量的现状,该研究提出了一种基于双模型渐进式的大口黑鲈投喂策略。采用单场多轮投喂方式,采集每轮投喂中包括音频在内的多模态数据。使用动态自适应阈值分割方法来检测单轮音频中大口黑鲈的摄食音频,基于提取的摄食音频特征进行主成分分析获取总摄食特征;使用特征选择算法筛选出大口黑鲈投喂状态的敏感特征作为模型的输入,用于学习和分类。结果表明,特征选择算法筛选出的大口黑鲈投喂状态的敏感特征为总摄食特征、光照强度、水温、体质量和单轮投喂率;对比了原始特征集与敏感特征集在梯度提升决策树和随机森林上的分类性能差异,结果证明了敏感特征的有效性;基于Transformer-RC的投喂状态预测模型在1~3时间窗口上的表现均优于其他三种模型,平均准确率为0.93~0.94;5种集成学习模型在投喂状态分类方面优于3种基础模型,其中基于LightGBM的投喂状态分类模型性能表现最佳,平均准确率为0.98;双模型渐进式投喂策略在饲料需求优化模型辅助下指导投喂,饵料系数较传统投喂降低约40%。该研究可为开发智能投喂系统提供依据。

, authors=李亚苹, 谭鹤群*, 陈以钒, 张义仁, authorsList=李亚苹, 谭鹤群, 陈以钒, 张义仁, authorCompany=null, correspAuthors=谭鹤群, authorNote=

李亚苹,博士生,研究方向为现代养殖技术与装备。Email:

, correspAuthorsNote=
谭鹤群,博士,教授,博士生导师,研究方向为畜禽水产养殖工程技术与装备。Email:
, copyrightStatement=版权所有 © 2026 农业工程学报编辑部, copyrightOwner=null, extLink=null, articleAbsUrl=null, sourceXml=tqi/8p/LKVbgl6CKSBRv7A==, magXml=pqmQKOWzlopYB32dRlhSWw==, pdfUrl=null, pdf=gE6vKWldiDL8QqG9n7HQQw==, pdfFileSize=2192808, pdfExtLink=null, richHtmlUrl=null, mobilePdfUrl=null, reviewReport=null, pdfFirstPage=null, abstractGraph=qPHBxMtRwahwKungaqN7FA==, abstractGraphContent=null, abstractVideo=null, citation=null, cebUrl=null, magXmlContent=9x2XGF+LWOl5oIK4o+kA7A==, mapNumber=null, fund=null)}, authors=[Author(id=1299828221348966659, tenantId=1146029695717560320, journalId=1296125453100220459, articleId=1297211746784141990, orderNo=0, firstName=null, middleName=null, lastName=null, nameCn=null, orcid=null, stid=null, country=null, authorPic=null, dead=0, email=liyaping@webmail.hzau.edu.cn, emailSecond=null, emailThird=null, correspondingAuthor=0, authorType=1, ext={EN=AuthorExt(id=1299828221424464134, tenantId=1146029695717560320, journalId=1296125453100220459, articleId=1297211746784141990, authorId=1299828221348966659, language=EN, stringName=Yaping LI, firstName=Yaping, middleName=null, lastName=LI, prefix=null, suffix=null, authorComment=null, nameInitials=null, affiliation=null, department=null, xref=null, address=1College of Engineering, Huazhong Agricultural University, Wuhan 430070, China
2The Key Laboratory of Aquaculture Facilities Engineering, Ministry of Agriculture and Rural Affairs, Wuhan 430070, China, bio=null, bioImg=null, bioContent=null, aboutCorrespAuthor=null), CN=AuthorExt(id=1299828221487378695, tenantId=1146029695717560320, journalId=1296125453100220459, articleId=1297211746784141990, authorId=1299828221348966659, language=CN, stringName=李亚苹, firstName=null, middleName=null, lastName=null, prefix=null, suffix=null, authorComment=null, nameInitials=null, affiliation=null, department=null, xref=null, address=1华中农业大学工学院,武汉 430070
2农业农村部水产养殖设施工程重点实验室,武汉 430070, bio={"content":"

李亚苹,博士生,研究方向为现代养殖技术与装备。Email:

"}, bioImg=null, bioContent=

李亚苹,博士生,研究方向为现代养殖技术与装备。Email:

, aboutCorrespAuthor=null)}, companyList=[AuthorCompany(id=1299828221177000188, tenantId=1146029695717560320, journalId=1296125453100220459, articleId=1297211746784141990, xref=1, ext=[AuthorCompanyExt(id=1299828221185388797, tenantId=1146029695717560320, journalId=1296125453100220459, articleId=1297211746784141990, companyId=1299828221177000188, language=EN, country=null, province=null, city=null, postcode=null, companyName=null, departmentName=null, remark=1College of Engineering, Huazhong Agricultural University, Wuhan 430070, China), AuthorCompanyExt(id=1299828221193777406, tenantId=1146029695717560320, journalId=1296125453100220459, articleId=1297211746784141990, companyId=1299828221177000188, language=CN, country=null, province=null, city=null, postcode=null, companyName=null, departmentName=null, remark=1华中农业大学工学院,武汉 430070)]), AuthorCompany(id=1299828221269274879, tenantId=1146029695717560320, journalId=1296125453100220459, articleId=1297211746784141990, xref=2, ext=[AuthorCompanyExt(id=1299828221277663488, tenantId=1146029695717560320, journalId=1296125453100220459, articleId=1297211746784141990, companyId=1299828221269274879, language=EN, country=null, province=null, city=null, postcode=null, companyName=null, departmentName=null, remark=2The Key Laboratory of Aquaculture Facilities Engineering, Ministry of Agriculture and Rural Affairs, Wuhan 430070, China), AuthorCompanyExt(id=1299828221281857793, tenantId=1146029695717560320, journalId=1296125453100220459, articleId=1297211746784141990, companyId=1299828221269274879, language=CN, country=null, province=null, city=null, postcode=null, companyName=null, departmentName=null, remark=2农业农村部水产养殖设施工程重点实验室,武汉 430070)])]), Author(id=1299828221579653385, tenantId=1146029695717560320, journalId=1296125453100220459, articleId=1297211746784141990, orderNo=1, firstName=null, middleName=null, lastName=null, nameCn=null, orcid=null, stid=null, country=null, authorPic=null, dead=0, email=thq@mail.hzau.edu.cn, emailSecond=null, emailThird=null, correspondingAuthor=1, authorType=1, ext={EN=AuthorExt(id=1299828221659345164, tenantId=1146029695717560320, journalId=1296125453100220459, articleId=1297211746784141990, authorId=1299828221579653385, language=EN, stringName=Hequn TAN, firstName=Hequn, middleName=null, lastName=TAN, prefix=null, suffix=null, authorComment=null, nameInitials=null, affiliation=null, department=null, xref=*, address=1College of Engineering, Huazhong Agricultural University, Wuhan 430070, China
2The Key Laboratory of Aquaculture Facilities Engineering, Ministry of Agriculture and Rural Affairs, Wuhan 430070, China, bio=null, bioImg=null, bioContent=null, aboutCorrespAuthor=null), CN=AuthorExt(id=1299828221734842637, tenantId=1146029695717560320, journalId=1296125453100220459, articleId=1297211746784141990, authorId=1299828221579653385, language=CN, stringName=谭鹤群, firstName=null, middleName=null, lastName=null, prefix=null, suffix=null, authorComment=null, nameInitials=null, affiliation=null, department=null, xref=*, address=1华中农业大学工学院,武汉 430070
2农业农村部水产养殖设施工程重点实验室,武汉 430070, bio=null, bioImg=null, bioContent=null, aboutCorrespAuthor=null)}, companyList=[AuthorCompany(id=1299828221177000188, tenantId=1146029695717560320, journalId=1296125453100220459, articleId=1297211746784141990, xref=1, ext=[AuthorCompanyExt(id=1299828221185388797, tenantId=1146029695717560320, journalId=1296125453100220459, articleId=1297211746784141990, companyId=1299828221177000188, language=EN, country=null, province=null, city=null, postcode=null, companyName=null, departmentName=null, remark=1College of Engineering, Huazhong Agricultural University, Wuhan 430070, China), AuthorCompanyExt(id=1299828221193777406, tenantId=1146029695717560320, journalId=1296125453100220459, articleId=1297211746784141990, companyId=1299828221177000188, language=CN, country=null, province=null, city=null, postcode=null, companyName=null, departmentName=null, remark=1华中农业大学工学院,武汉 430070)]), AuthorCompany(id=1299828221269274879, tenantId=1146029695717560320, journalId=1296125453100220459, articleId=1297211746784141990, xref=2, ext=[AuthorCompanyExt(id=1299828221277663488, tenantId=1146029695717560320, journalId=1296125453100220459, articleId=1297211746784141990, companyId=1299828221269274879, language=EN, country=null, province=null, city=null, postcode=null, companyName=null, departmentName=null, remark=2The Key Laboratory of Aquaculture Facilities Engineering, Ministry of Agriculture and Rural Affairs, Wuhan 430070, China), AuthorCompanyExt(id=1299828221281857793, tenantId=1146029695717560320, journalId=1296125453100220459, articleId=1297211746784141990, companyId=1299828221269274879, language=CN, country=null, province=null, city=null, postcode=null, companyName=null, departmentName=null, remark=2农业农村部水产养殖设施工程重点实验室,武汉 430070)])]), Author(id=1299828221814534415, tenantId=1146029695717560320, journalId=1296125453100220459, articleId=1297211746784141990, orderNo=2, firstName=null, middleName=null, lastName=null, nameCn=null, orcid=null, stid=null, country=null, authorPic=null, dead=0, email=null, emailSecond=null, emailThird=null, correspondingAuthor=0, authorType=1, ext={EN=AuthorExt(id=1299828221902614802, tenantId=1146029695717560320, journalId=1296125453100220459, articleId=1297211746784141990, authorId=1299828221814534415, language=EN, stringName=Yifan CHEN, firstName=Yifan, middleName=null, lastName=CHEN, prefix=null, suffix=null, authorComment=null, nameInitials=null, affiliation=null, department=null, xref=null, address=1College of Engineering, Huazhong Agricultural University, Wuhan 430070, China
2The Key Laboratory of Aquaculture Facilities Engineering, Ministry of Agriculture and Rural Affairs, Wuhan 430070, China, bio=null, bioImg=null, bioContent=null, aboutCorrespAuthor=null), CN=AuthorExt(id=1299828221986500883, tenantId=1146029695717560320, journalId=1296125453100220459, articleId=1297211746784141990, authorId=1299828221814534415, language=CN, stringName=陈以钒, firstName=null, middleName=null, lastName=null, prefix=null, suffix=null, authorComment=null, nameInitials=null, affiliation=null, department=null, xref=null, address=1华中农业大学工学院,武汉 430070
2农业农村部水产养殖设施工程重点实验室,武汉 430070, bio=null, bioImg=null, bioContent=null, aboutCorrespAuthor=null)}, companyList=[AuthorCompany(id=1299828221177000188, tenantId=1146029695717560320, journalId=1296125453100220459, articleId=1297211746784141990, xref=1, ext=[AuthorCompanyExt(id=1299828221185388797, tenantId=1146029695717560320, journalId=1296125453100220459, articleId=1297211746784141990, companyId=1299828221177000188, language=EN, country=null, province=null, city=null, postcode=null, companyName=null, departmentName=null, remark=1College of Engineering, Huazhong Agricultural University, Wuhan 430070, China), AuthorCompanyExt(id=1299828221193777406, tenantId=1146029695717560320, journalId=1296125453100220459, articleId=1297211746784141990, companyId=1299828221177000188, language=CN, country=null, province=null, city=null, postcode=null, companyName=null, departmentName=null, remark=1华中农业大学工学院,武汉 430070)]), AuthorCompany(id=1299828221269274879, tenantId=1146029695717560320, journalId=1296125453100220459, articleId=1297211746784141990, xref=2, ext=[AuthorCompanyExt(id=1299828221277663488, tenantId=1146029695717560320, journalId=1296125453100220459, articleId=1297211746784141990, companyId=1299828221269274879, language=EN, country=null, province=null, city=null, postcode=null, companyName=null, departmentName=null, remark=2The Key Laboratory of Aquaculture Facilities Engineering, Ministry of Agriculture and Rural Affairs, Wuhan 430070, China), AuthorCompanyExt(id=1299828221281857793, tenantId=1146029695717560320, journalId=1296125453100220459, articleId=1297211746784141990, companyId=1299828221269274879, language=CN, country=null, province=null, city=null, postcode=null, companyName=null, departmentName=null, remark=2农业农村部水产养殖设施工程重点实验室,武汉 430070)])]), Author(id=1299828222036832533, tenantId=1146029695717560320, journalId=1296125453100220459, articleId=1297211746784141990, orderNo=3, firstName=null, middleName=null, lastName=null, nameCn=null, orcid=null, stid=null, country=null, authorPic=null, dead=0, email=null, emailSecond=null, emailThird=null, correspondingAuthor=0, authorType=1, ext={EN=AuthorExt(id=1299828222124912920, tenantId=1146029695717560320, journalId=1296125453100220459, articleId=1297211746784141990, authorId=1299828222036832533, language=EN, stringName=Yiren ZHANG, firstName=Yiren, middleName=null, lastName=ZHANG, prefix=null, suffix=null, authorComment=null, nameInitials=null, affiliation=null, department=null, xref=null, address=1College of Engineering, Huazhong Agricultural University, Wuhan 430070, China
2The Key Laboratory of Aquaculture Facilities Engineering, Ministry of Agriculture and Rural Affairs, Wuhan 430070, China, bio=null, bioImg=null, bioContent=null, aboutCorrespAuthor=null), CN=AuthorExt(id=1299828222196216089, tenantId=1146029695717560320, journalId=1296125453100220459, articleId=1297211746784141990, authorId=1299828222036832533, language=CN, stringName=张义仁, firstName=null, middleName=null, lastName=null, prefix=null, suffix=null, authorComment=null, nameInitials=null, affiliation=null, department=null, xref=null, address=1华中农业大学工学院,武汉 430070
2农业农村部水产养殖设施工程重点实验室,武汉 430070, bio=null, bioImg=null, bioContent=null, aboutCorrespAuthor=null)}, companyList=[AuthorCompany(id=1299828221177000188, tenantId=1146029695717560320, journalId=1296125453100220459, articleId=1297211746784141990, xref=1, ext=[AuthorCompanyExt(id=1299828221185388797, tenantId=1146029695717560320, journalId=1296125453100220459, articleId=1297211746784141990, companyId=1299828221177000188, language=EN, country=null, province=null, city=null, postcode=null, companyName=null, departmentName=null, remark=1College of Engineering, Huazhong Agricultural University, Wuhan 430070, China), AuthorCompanyExt(id=1299828221193777406, tenantId=1146029695717560320, journalId=1296125453100220459, articleId=1297211746784141990, companyId=1299828221177000188, language=CN, country=null, province=null, city=null, postcode=null, companyName=null, departmentName=null, remark=1华中农业大学工学院,武汉 430070)]), AuthorCompany(id=1299828221269274879, tenantId=1146029695717560320, journalId=1296125453100220459, articleId=1297211746784141990, xref=2, ext=[AuthorCompanyExt(id=1299828221277663488, tenantId=1146029695717560320, journalId=1296125453100220459, articleId=1297211746784141990, companyId=1299828221269274879, language=EN, country=null, province=null, city=null, postcode=null, companyName=null, departmentName=null, remark=2The Key Laboratory of Aquaculture Facilities Engineering, Ministry of Agriculture and Rural Affairs, Wuhan 430070, China), AuthorCompanyExt(id=1299828221281857793, tenantId=1146029695717560320, journalId=1296125453100220459, articleId=1297211746784141990, companyId=1299828221269274879, language=CN, country=null, province=null, city=null, postcode=null, companyName=null, departmentName=null, remark=2农业农村部水产养殖设施工程重点实验室,武汉 430070)])])], keywords=[Keyword(id=1299828222334628122, tenantId=1146029695717560320, journalId=1296125453100220459, articleId=1297211746784141990, language=EN, orderNo=1, keyword=aquaculture), Keyword(id=1299828222397542683, tenantId=1146029695717560320, journalId=1296125453100220459, articleId=1297211746784141990, language=EN, orderNo=2, keyword=Micropterus salmoides), Keyword(id=1299828222464651548, tenantId=1146029695717560320, journalId=1296125453100220459, articleId=1297211746784141990, language=EN, orderNo=3, keyword=precision feeding), Keyword(id=1299828222531760413, tenantId=1146029695717560320, journalId=1296125453100220459, articleId=1297211746784141990, language=EN, orderNo=4, keyword=dual-model), Keyword(id=1299828222611452190, tenantId=1146029695717560320, journalId=1296125453100220459, articleId=1297211746784141990, language=EN, orderNo=5, keyword=feeding status), Keyword(id=1299828222691143967, tenantId=1146029695717560320, journalId=1296125453100220459, articleId=1297211746784141990, language=CN, orderNo=1, keyword=水产养殖), Keyword(id=1299828222770835744, tenantId=1146029695717560320, journalId=1296125453100220459, articleId=1297211746784141990, language=CN, orderNo=2, keyword=大口黑鲈), Keyword(id=1299828222846333217, tenantId=1146029695717560320, journalId=1296125453100220459, articleId=1297211746784141990, language=CN, orderNo=3, keyword=精准投喂), Keyword(id=1299828222921830690, tenantId=1146029695717560320, journalId=1296125453100220459, articleId=1297211746784141990, language=CN, orderNo=4, keyword=双模型), Keyword(id=1299828222993133859, tenantId=1146029695717560320, journalId=1296125453100220459, articleId=1297211746784141990, language=CN, orderNo=5, keyword=投喂策略)], refs=[Reference(id=1299828226159833420, tenantId=1146029695717560320, journalId=1296125453100220459, articleId=1297211746784141990, doi=null, pmid=null, pmcid=null, year=null, volume=null, issue=null, pageStart=null, pageEnd=null, url=null, language=null, rfNumber=1, rfOrder=0, authorNames=null, journalName=null, refType=null, unstructuredReference=LI D L, WANG Z H, WU S Y, et al. Automatic recognition methods of fish feeding behavior in aquaculture: A review[J]. Aquaculture, 2020, 528: 735508., articleTitle=null, refAbstract=null), Reference(id=1299828226222747981, tenantId=1146029695717560320, journalId=1296125453100220459, articleId=1297211746784141990, doi=null, pmid=null, pmcid=null, year=null, volume=null, issue=null, pageStart=null, pageEnd=null, url=null, language=null, rfNumber=2, rfOrder=1, authorNames=null, journalName=null, refType=null, unstructuredReference=WU T H, HUANG Y I, CHEN J M. Development of an adaptive neural-based fuzzy inference system for feeding decision-making assessment in silver perch (Bidyanus bidyanus) culture[J]. Aquacultural Engineering, 2015, 66: 41-51., articleTitle=null, refAbstract=null), Reference(id=1299828226306634062, tenantId=1146029695717560320, journalId=1296125453100220459, articleId=1297211746784141990, doi=null, pmid=null, pmcid=null, year=null, volume=null, issue=null, pageStart=null, pageEnd=null, url=null, language=null, rfNumber=3, rfOrder=2, authorNames=null, journalName=null, refType=null, unstructuredReference=ZHOU C, ZHANG B H, LIN K, et al. Near-infrared imaging to quantify the feeding behavior of fish in aquaculture[J]. Computers and Electronics in Agriculture, 2017, 135(C): 233-241., articleTitle=null, refAbstract=null), Reference(id=1299828226365354319, tenantId=1146029695717560320, journalId=1296125453100220459, articleId=1297211746784141990, doi=null, pmid=null, pmcid=null, year=null, volume=null, issue=null, pageStart=null, pageEnd=null, url=null, language=null, rfNumber=4, rfOrder=3, authorNames=null, journalName=null, refType=null, unstructuredReference=BARRAZA-GUARDADO R H, MARTÍNEZ-CÓRDOVA L R, ENRÍQUEZ-OCAñA L F, et al. Effect of shrimp farm effluent on water and sediment quality parameters off the coast of Sonora, México[J]. Ciencias Marinas, 2014, 40(4): 221-235., articleTitle=null, refAbstract=null), Reference(id=1299828226445046096, tenantId=1146029695717560320, journalId=1296125453100220459, articleId=1297211746784141990, doi=null, pmid=null, pmcid=null, year=null, volume=null, issue=null, pageStart=null, pageEnd=null, url=null, language=null, rfNumber=5, rfOrder=4, authorNames=null, journalName=null, refType=null, unstructuredReference=ZHOU C, YANG X T, ZHANG B H, et al. An adaptive image enhancement method for a recirculating aquaculture system[J]. Scientific Reports, 2017, 7: 6243., articleTitle=null, refAbstract=null), Reference(id=1299828226528932177, tenantId=1146029695717560320, journalId=1296125453100220459, articleId=1297211746784141990, doi=null, pmid=null, pmcid=null, year=null, volume=null, issue=null, pageStart=null, pageEnd=null, url=null, language=null, rfNumber=6, rfOrder=5, authorNames=null, journalName=null, refType=null, unstructuredReference=LIU Z, LI, X, FAN L Z, et al. Measuring feeding activity of fish in RAS using computer vision[J]. Aquacultural Engineering, 2014, 60: 20-27., articleTitle=null, refAbstract=null), Reference(id=1299828226604429650, tenantId=1146029695717560320, journalId=1296125453100220459, articleId=1297211746784141990, doi=null, pmid=null, pmcid=null, year=null, volume=null, issue=null, pageStart=null, pageEnd=null, url=null, language=null, rfNumber=7, rfOrder=6, authorNames=null, journalName=null, refType=null, unstructuredReference=ZHOU C, LIN K, XU D M, et al. Near infrared computer vision and neuro-fuzzy model-based feeding decision system for fish in aquaculture[J]. Computers and Electronics in Agriculture, 2018, 146: 114-124., articleTitle=null, refAbstract=null), Reference(id=1299828226667344211, tenantId=1146029695717560320, journalId=1296125453100220459, articleId=1297211746784141990, doi=null, pmid=null, pmcid=null, year=null, volume=null, issue=null, pageStart=null, pageEnd=null, url=null, language=null, rfNumber=8, rfOrder=7, authorNames=null, journalName=null, refType=null, unstructuredReference=ZHOU C, XU D M, LIN K, et al. Intelligent feeding control methods in aquaculture with an emphasis on fish: A review[J]. Reviews in Aquaculture, 2017, 10(4): 975-993., articleTitle=null, refAbstract=null), Reference(id=1299828226734453076, tenantId=1146029695717560320, journalId=1296125453100220459, articleId=1297211746784141990, doi=null, pmid=null, pmcid=null, year=null, volume=null, issue=null, pageStart=null, pageEnd=null, url=null, language=null, rfNumber=9, rfOrder=8, authorNames=null, journalName=null, refType=null, unstructuredReference=DAVIS D A, HARDY R H. Feeding and Fish Husbandry[M]. London: Academic Press, 2022., articleTitle=null, refAbstract=null), Reference(id=1299828226793173333, tenantId=1146029695717560320, journalId=1296125453100220459, articleId=1297211746784141990, doi=null, pmid=null, pmcid=null, year=null, volume=null, issue=null, pageStart=null, pageEnd=null, url=null, language=null, rfNumber=10, rfOrder=9, authorNames=null, journalName=null, refType=null, unstructuredReference=LIU X J, DU K H, ZHANG C Y, et al. Precision feeding system for largemouth bass (Micropterus salmoides) based on multi-factor comprehensive control[J]. Biosystems Engineering, 2023, 227: 195-216., articleTitle=null, refAbstract=null), Reference(id=1299828226868670806, tenantId=1146029695717560320, journalId=1296125453100220459, articleId=1297211746784141990, doi=null, pmid=null, pmcid=null, year=null, volume=null, issue=null, pageStart=null, pageEnd=null, url=null, language=null, rfNumber=11, rfOrder=10, authorNames=null, journalName=null, refType=null, unstructuredReference=MA P C, YANG X T, HU W C, et al. Fish feeding behavior recognition using time-domain and frequency-domain signals fusion from six-axis inertial sensors[J]. Computers and Electronics in Agriculture, 2024, 227(2): 109652., articleTitle=null, refAbstract=null), Reference(id=1299828226965139799, tenantId=1146029695717560320, journalId=1296125453100220459, articleId=1297211746784141990, doi=null, pmid=null, pmcid=null, year=null, volume=null, issue=null, pageStart=null, pageEnd=null, url=null, language=null, rfNumber=12, rfOrder=11, authorNames=null, journalName=null, refType=null, unstructuredReference=STONER A W, OTTMAR M L, HURST T P. Temperature affects activity and feeding motivation in Pacific halibut: Implications for bait-dependent fishing[J]. Fisheries Research, 2006, 81(2-3): 202-209., articleTitle=null, refAbstract=null), Reference(id=1299828227065803096, tenantId=1146029695717560320, journalId=1296125453100220459, articleId=1297211746784141990, doi=null, pmid=null, pmcid=null, year=null, volume=null, issue=null, pageStart=null, pageEnd=null, url=null, language=null, rfNumber=13, rfOrder=12, authorNames=null, journalName=null, refType=null, unstructuredReference=CHEN L, YANG XT, SUN CH, et al. Feed intake prediction model for group fish using the MEA-BP neural network in intensive aquaculture[J]. Information Processing in Agriculture, 2020, 7(2): 261-271., articleTitle=null, refAbstract=null), Reference(id=1299828227141300569, tenantId=1146029695717560320, journalId=1296125453100220459, articleId=1297211746784141990, doi=null, pmid=null, pmcid=null, year=null, volume=null, issue=null, pageStart=null, pageEnd=null, url=null, language=null, rfNumber=14, rfOrder=13, authorNames=null, journalName=null, refType=null, unstructuredReference=UBINA N M, CHENG S-C, CHANG C-C, et al. Evaluating fish feeding intensity in aquaculture with convolutional neural networks[J]. Aquacultural Engineering, 2021, 94: 102178., articleTitle=null, refAbstract=null), Reference(id=1299828227220992346, tenantId=1146029695717560320, journalId=1296125453100220459, articleId=1297211746784141990, doi=null, pmid=null, pmcid=null, year=null, volume=null, issue=null, pageStart=null, pageEnd=null, url=null, language=null, rfNumber=15, rfOrder=14, authorNames=null, journalName=null, refType=null, unstructuredReference=WANG Y Q, YU X N, LIU J C, et al. Dynamic feeding method for aquaculture fish using multi-task neural network[J]. Aquaculture, 2022, 551: 737913., articleTitle=null, refAbstract=null), Reference(id=1299828227288101211, tenantId=1146029695717560320, journalId=1296125453100220459, articleId=1297211746784141990, doi=null, pmid=null, pmcid=null, year=null, volume=null, issue=null, pageStart=null, pageEnd=null, url=null, language=null, rfNumber=16, rfOrder=15, authorNames=null, journalName=null, refType=null, unstructuredReference=CAI K W, YANG Z P, GAO T Y, et al. Efficient recognition of fish feeding behavior: A novel two-stage framework pioneering intelligent aquaculture strategies[J]. Computers and Electronics in Agriculture, 2024, 224: 109129., articleTitle=null, refAbstract=null), Reference(id=1299828227363598684, tenantId=1146029695717560320, journalId=1296125453100220459, articleId=1297211746784141990, doi=null, pmid=null, pmcid=null, year=null, volume=null, issue=null, pageStart=null, pageEnd=null, url=null, language=null, rfNumber=17, rfOrder=16, authorNames=null, journalName=null, refType=null, unstructuredReference=LIU H Y, XU L H, LI D W. Detection and recognition of uneaten fish food pellets in aquaculture using image processing[C]. Sixth International Conference on Graphic and Image Processing (ICGIP), 2015, 9443: 94430G., articleTitle=null, refAbstract=null), Reference(id=1299828227430707549, tenantId=1146029695717560320, journalId=1296125453100220459, articleId=1297211746784141990, doi=null, pmid=null, pmcid=null, year=null, volume=null, issue=null, pageStart=null, pageEnd=null, url=null, language=null, rfNumber=18, rfOrder=17, authorNames=null, journalName=null, refType=null, unstructuredReference=LI D W, XU L H, LIU H Y. Detection of uneaten fish food pellets in underwater images for aquaculture[J]. Aquacultural Engineering, 2017, 78(Part B): 85-94., articleTitle=null, refAbstract=null), Reference(id=1299828227497816414, tenantId=1146029695717560320, journalId=1296125453100220459, articleId=1297211746784141990, doi=null, pmid=null, pmcid=null, year=null, volume=null, issue=null, pageStart=null, pageEnd=null, url=null, language=null, rfNumber=19, rfOrder=18, authorNames=null, journalName=null, refType=null, unstructuredReference=HUNG C C, TSAO S C, HUANG K H, et al. A highly sensitive underwater video system for use in turbid aquaculture ponds[J]. Scientific Reports, 2016, 6: 31810., articleTitle=null, refAbstract=null), Reference(id=1299828227560730975, tenantId=1146029695717560320, journalId=1296125453100220459, articleId=1297211746784141990, doi=null, pmid=null, pmcid=null, year=null, volume=null, issue=null, pageStart=null, pageEnd=null, url=null, language=null, rfNumber=20, rfOrder=19, authorNames=null, journalName=null, refType=null, unstructuredReference=ANRAS M B, LAGARDÈRE J P. Measuring cultured fish swimming behaviour: first results on rainbow trout using acoustic telemetry in tanks[J]. Aquaculture, 2004, 240(1/2/3/4): 175-186., articleTitle=null, refAbstract=null), Reference(id=1299828227619451232, tenantId=1146029695717560320, journalId=1296125453100220459, articleId=1297211746784141990, doi=null, pmid=null, pmcid=null, year=null, volume=null, issue=null, pageStart=null, pageEnd=null, url=null, language=null, rfNumber=21, rfOrder=20, authorNames=null, journalName=null, refType=null, unstructuredReference=DU Z Z, CUI M, WANG Q, et al. Feeding intensity assessment of aquaculture fish using Mel Spectrogram and deep learning algorithms[J]. Aquacultural Engineering, 2023, 102: 102345., articleTitle=null, refAbstract=null), Reference(id=1299828227694948705, tenantId=1146029695717560320, journalId=1296125453100220459, articleId=1297211746784141990, doi=null, pmid=null, pmcid=null, year=null, volume=null, issue=null, pageStart=null, pageEnd=null, url=null, language=null, rfNumber=22, rfOrder=21, authorNames=null, journalName=null, refType=null, unstructuredReference=ZENG Y H, YANG X T, PAN L, et al. Fish school feeding behavior quantification using acoustic signal and improved Swin Transformer[J]. Computers and Electronics in Agriculture, 2023, 204: 107580., articleTitle=null, refAbstract=null), Reference(id=1299828227770446178, tenantId=1146029695717560320, journalId=1296125453100220459, articleId=1297211746784141990, doi=null, pmid=null, pmcid=null, year=null, volume=null, issue=null, pageStart=null, pageEnd=null, url=null, language=null, rfNumber=23, rfOrder=22, authorNames=null, journalName=null, refType=null, unstructuredReference=CAO X K, REN N, TIAN G L, et al. A three-dimensional prediction method of dissolved oxygen in pond culture based on Attention-GRU-GBRT[J]. Computers and Electronics in Agriculture, 2021, 181: 105955., articleTitle=null, refAbstract=null), Reference(id=1299828227845943651, tenantId=1146029695717560320, journalId=1296125453100220459, articleId=1297211746784141990, doi=null, pmid=null, pmcid=null, year=null, volume=null, issue=null, pageStart=null, pageEnd=null, url=null, language=null, rfNumber=24, rfOrder=23, authorNames=null, journalName=null, refType=null, unstructuredReference=GUO J J, DONG J Q, ZHOU B, et al. A hybrid model for the prediction of dissolved oxygen in seabass farming[J]. Computers and Electronics in Agriculture, 2022, 198: 106971., articleTitle=null, refAbstract=null), Reference(id=1299828227946606948, tenantId=1146029695717560320, journalId=1296125453100220459, articleId=1297211746784141990, doi=null, pmid=null, pmcid=null, year=null, volume=null, issue=null, pageStart=null, pageEnd=null, url=null, language=null, rfNumber=25, rfOrder=24, authorNames=null, journalName=null, refType=null, unstructuredReference=XU C, DING J Q, QIAO Y, et al. Tomato disease and pest diagnosis method based on the Stacking of prescription data[J]. Computers and Electronics in Agriculture, 2022, 197: 106997., articleTitle=null, refAbstract=null), Reference(id=1299828228030493029, tenantId=1146029695717560320, journalId=1296125453100220459, articleId=1297211746784141990, doi=null, pmid=null, pmcid=null, year=null, volume=null, issue=null, pageStart=null, pageEnd=null, url=null, language=null, rfNumber=26, rfOrder=25, authorNames=null, journalName=null, refType=null, unstructuredReference=JIANG H W, ZHANG S L, YANG Z, et al. Quality classification of stored wheat based on evidence reasoning rule and stacking ensemble learning[J]. Computers and Electronics in Agriculture, 2023, 214: 108339., articleTitle=null, refAbstract=null), Reference(id=1299828228122767718, tenantId=1146029695717560320, journalId=1296125453100220459, articleId=1297211746784141990, doi=null, pmid=null, pmcid=null, year=null, volume=null, issue=null, pageStart=null, pageEnd=null, url=null, language=null, rfNumber=27, rfOrder=26, authorNames=null, journalName=null, refType=null, unstructuredReference=ØVERLI Ø, SøRENSEN C, NILSSON G E. Behavioral indicators of stress-coping style in rainbow trout: Do males and females react differently to novelty?[J]. Physiology & Behavior, 2006, 87(3): 506-512., articleTitle=null, refAbstract=null), Reference(id=1299828228189876584, tenantId=1146029695717560320, journalId=1296125453100220459, articleId=1297211746784141990, doi=null, pmid=null, pmcid=null, year=null, volume=null, issue=null, pageStart=null, pageEnd=null, url=null, language=null, rfNumber=28, rfOrder=27, authorNames=null, journalName=null, refType=null, unstructuredReference=UPADHYAY N, KARMAKAR A. An improved multi-band spectral subtraction algorithm for enhancing speech in various noise environments[J]. Procedia Engineering, 2013, 64: 312-321., articleTitle=null, refAbstract=null), Reference(id=1299828228236013929, tenantId=1146029695717560320, journalId=1296125453100220459, articleId=1297211746784141990, doi=null, pmid=null, pmcid=null, year=null, volume=null, issue=null, pageStart=null, pageEnd=null, url=null, language=null, rfNumber=29, rfOrder=28, authorNames=null, journalName=null, refType=null, unstructuredReference=BEROUTI M, SCHWARTZ R, MAKHOUL J. Enhancement of speech corrupted by acoustic noise[C]. IEEE International Conference on Acoustics, Speech, and Signal Processing (ICASSP). Washington, USA: IEEE, 1979: 208-211., articleTitle=null, refAbstract=null), Reference(id=1299828228307317098, tenantId=1146029695717560320, journalId=1296125453100220459, articleId=1297211746784141990, doi=null, pmid=null, pmcid=null, year=null, volume=null, issue=null, pageStart=null, pageEnd=null, url=null, language=null, rfNumber=30, rfOrder=29, authorNames=null, journalName=null, refType=null, unstructuredReference=曹晓慧, 刘晃, 戚仁宇, 等. 循环水养殖大口黑鲈摄食颗粒饲料的声学特征[J]. 农业工程学报, 2021, 37(20): 219-225., articleTitle=null, refAbstract=null), Reference(id=1299828228370231659, tenantId=1146029695717560320, journalId=1296125453100220459, articleId=1297211746784141990, doi=null, pmid=null, pmcid=null, year=null, volume=null, issue=null, pageStart=null, pageEnd=null, url=null, language=null, rfNumber=30, rfOrder=30, authorNames=null, journalName=null, refType=null, unstructuredReference=CAO Xiaohui, LIU Huang, QI Renyu, et al. Acoustic characteristics of the feeding pellets for Micropterus salmoides in circulating aquaculture[J]. Transactions of the Chinese Society of Agricultural Engineering (Transactions of the CSAE), 2021, 37(20): 219-225. (in Chinese with English abstract), articleTitle=null, refAbstract=null), Reference(id=1299828228433146220, tenantId=1146029695717560320, journalId=1296125453100220459, articleId=1297211746784141990, doi=null, pmid=null, pmcid=null, year=null, volume=null, issue=null, pageStart=null, pageEnd=null, url=null, language=null, rfNumber=31, rfOrder=31, authorNames=null, journalName=null, refType=null, unstructuredReference=LI Y P, TAN H Q, ZHANG Y R, et al. A quantitative method of feeding demand for M. salmoides based on multi-source information fusion[J]. Biosystems Engineering, 2026, 262: 104372., articleTitle=null, refAbstract=null), Reference(id=1299828228496060781, tenantId=1146029695717560320, journalId=1296125453100220459, articleId=1297211746784141990, doi=null, pmid=null, pmcid=null, year=null, volume=null, issue=null, pageStart=null, pageEnd=null, url=null, language=null, rfNumber=32, rfOrder=32, authorNames=null, journalName=null, refType=null, unstructuredReference=MALLEKH R, LAGARDÈRE J P, ENEAU J P, et al. An acoustic detector of turbot feeding activity[J]. Aquaculture, 2003, 221(1/2/3/4): 481-489., articleTitle=null, refAbstract=null), Reference(id=1299828228563169646, tenantId=1146029695717560320, journalId=1296125453100220459, articleId=1297211746784141990, doi=null, pmid=null, pmcid=null, year=null, volume=null, issue=null, pageStart=null, pageEnd=null, url=null, language=null, rfNumber=33, rfOrder=33, authorNames=null, journalName=null, refType=null, unstructuredReference=KAVEH M, SHARABIANI V R, CHAYJAN R A, et al. ANFIS and ANNs model for prediction of moisture diffusivity and specific energy consumption potato, garlic and cantaloupe drying under convective hot air dryer[J]. Information Processing in Agriculture, 2018, 5(3): 372-387., articleTitle=null, refAbstract=null), Reference(id=1299828228638667119, tenantId=1146029695717560320, journalId=1296125453100220459, articleId=1297211746784141990, doi=null, pmid=null, pmcid=null, year=null, volume=null, issue=null, pageStart=null, pageEnd=null, url=null, language=null, rfNumber=34, rfOrder=34, authorNames=null, journalName=null, refType=null, unstructuredReference=DELGADO-PANADERO A, HERNÁNDEZ-LORCA B, GARCÍA-ORDÁS M. T, et al. Implementing local-explainability in gradient boosting trees: Feature contribution[J]. Information Sciences, 2022, 589: 199-212., articleTitle=null, refAbstract=null), Reference(id=1299828228709970288, tenantId=1146029695717560320, journalId=1296125453100220459, articleId=1297211746784141990, doi=null, pmid=null, pmcid=null, year=null, volume=null, issue=null, pageStart=null, pageEnd=null, url=null, language=null, rfNumber=35, rfOrder=35, authorNames=null, journalName=null, refType=null, unstructuredReference=DHILLON R, TAKOO G, SHARMA V, et al. Utilizing machine learning framework to evaluate the effect of climate change on maize and soybean yield[J]. Computers and Electronics in Agriculture, 2024, 221: 108982., articleTitle=null, refAbstract=null), Reference(id=1299828228789662065, tenantId=1146029695717560320, journalId=1296125453100220459, articleId=1297211746784141990, doi=null, pmid=null, pmcid=null, year=null, volume=null, issue=null, pageStart=null, pageEnd=null, url=null, language=null, rfNumber=36, rfOrder=36, authorNames=null, journalName=null, refType=null, unstructuredReference=KIRANYAZ S, INCE T, ABDELJABER O, et al. 1-D Convolutional Neural Networks for Signal Processing Applications[C]// IEEE International Conference on Acoustics, Speech and Signal Processing (ICASSP), Brighton, UK: IEEE, 2019, 8360-8364., articleTitle=null, refAbstract=null), Reference(id=1299828228856770930, tenantId=1146029695717560320, journalId=1296125453100220459, articleId=1297211746784141990, doi=null, pmid=null, pmcid=null, year=null, volume=null, issue=null, pageStart=null, pageEnd=null, url=null, language=null, rfNumber=37, rfOrder=37, authorNames=null, journalName=null, refType=null, unstructuredReference=MENG J Q, LI C B, TAO J, et al. RNN- LSTM-based model predictive control for a corn-to-sugar process[J]. Processes, 2023, 11(4): 1080., articleTitle=null, refAbstract=null), Reference(id=1299828228923879795, tenantId=1146029695717560320, journalId=1296125453100220459, articleId=1297211746784141990, doi=null, pmid=null, pmcid=null, year=null, volume=null, issue=null, pageStart=null, pageEnd=null, url=null, language=null, rfNumber=38, rfOrder=38, authorNames=null, journalName=null, refType=null, unstructuredReference=李雯, 施珮. 融合特征信息和改进BiGRU的池塘养殖溶解氧预测模型[J]. 农业工程学报, 2025, 41(19): 218-226., articleTitle=null, refAbstract=null), Reference(id=1299828228995182964, tenantId=1146029695717560320, journalId=1296125453100220459, articleId=1297211746784141990, doi=null, pmid=null, pmcid=null, year=null, volume=null, issue=null, pageStart=null, pageEnd=null, url=null, language=null, rfNumber=38, rfOrder=39, authorNames=null, journalName=null, refType=null, unstructuredReference=LI Wen, SHI Pei. Dissolved oxygen prediction model for pond culture by integrating feature information and improved BiGRU[J]. Transactions of the Chinese Society of Agricultural Engineering (Transactions of the CSAE), 2025, 41(19): 218-226. (in Chinese with English abstract), articleTitle=null, refAbstract=null), Reference(id=1299828229053903221, tenantId=1146029695717560320, journalId=1296125453100220459, articleId=1297211746784141990, doi=null, pmid=null, pmcid=null, year=null, volume=null, issue=null, pageStart=null, pageEnd=null, url=null, language=null, rfNumber=39, rfOrder=40, authorNames=null, journalName=null, refType=null, unstructuredReference=LI Y X, JIA W X, WANG J M, et al. ALeRT-COVID: attentive lockdown-awaRe transfer learning for predicting COVID-19 pandemics in different countries[J]. Journal of Healthcare Informatics Research, 2021, 5: 98-113., articleTitle=null, refAbstract=null), Reference(id=1299828229137789302, tenantId=1146029695717560320, journalId=1296125453100220459, articleId=1297211746784141990, doi=null, pmid=null, pmcid=null, year=null, volume=null, issue=null, pageStart=null, pageEnd=null, url=null, language=null, rfNumber=40, rfOrder=41, authorNames=null, journalName=null, refType=null, unstructuredReference=LIU Y Q, ZHANG Q, SONG L H, et al. Attention-based recurrent neural networks for accurate short-term and long- term dissolved oxygen prediction[J]. Computers and Electronics in Agriculture, 2019, 165: 104964., articleTitle=null, refAbstract=null), Reference(id=1299828229213286775, tenantId=1146029695717560320, journalId=1296125453100220459, articleId=1297211746784141990, doi=null, pmid=null, pmcid=null, year=null, volume=null, issue=null, pageStart=null, pageEnd=null, url=null, language=null, rfNumber=41, rfOrder=42, authorNames=null, journalName=null, refType=null, unstructuredReference=COVER T, HART P. Nearest neighbor pattern classification[J]. IEEE Transactions on Information Theory, 1967, 13(1): 21-27., articleTitle=null, refAbstract=null), Reference(id=1299828229288784248, tenantId=1146029695717560320, journalId=1296125453100220459, articleId=1297211746784141990, doi=null, pmid=null, pmcid=null, year=null, volume=null, issue=null, pageStart=null, pageEnd=null, url=null, language=null, rfNumber=42, rfOrder=43, authorNames=null, journalName=null, refType=null, unstructuredReference=BOSER B E, GUYON I, VAPNIK V N. A training algorithm for optimal margin classifiers[C]// Proceedings of the fifth annual workshop on Computational learning theory. New York, USA: Association for Computing Machinery, 1992, 144-152., articleTitle=null, refAbstract=null), Reference(id=1299828229368476025, tenantId=1146029695717560320, journalId=1296125453100220459, articleId=1297211746784141990, doi=null, pmid=null, pmcid=null, year=null, volume=null, issue=null, pageStart=null, pageEnd=null, url=null, language=null, rfNumber=43, rfOrder=44, authorNames=null, journalName=null, refType=null, unstructuredReference=BREIMAN L. Random forests[J]. Machine Learning, 2001, 45: 5-32., articleTitle=null, refAbstract=null), Reference(id=1299828229443973498, tenantId=1146029695717560320, journalId=1296125453100220459, articleId=1297211746784141990, doi=null, pmid=null, pmcid=null, year=null, volume=null, issue=null, pageStart=null, pageEnd=null, url=null, language=null, rfNumber=44, rfOrder=45, authorNames=null, journalName=null, refType=null, unstructuredReference=FREUND Y, SCHAPIRE R E. Experiment with a New Boosting Algorithm[C]// Proceedings of the Thirteenth International Conference on Machine Learning. San Francisco, CA, USA: Morgan Kaufmann Publishers Inc. , 1996, 148-156., articleTitle=null, refAbstract=null), Reference(id=1299828229515276667, tenantId=1146029695717560320, journalId=1296125453100220459, articleId=1297211746784141990, doi=null, pmid=null, pmcid=null, year=null, volume=null, issue=null, pageStart=null, pageEnd=null, url=null, language=null, rfNumber=45, rfOrder=46, authorNames=null, journalName=null, refType=null, unstructuredReference=FRIEDMAN J H. Greedy Function Approximation: A gradient boosting machine[J]. Annals of Statistics, 2001, 29: 1189-1232., articleTitle=null, refAbstract=null), Reference(id=1299828229590774140, tenantId=1146029695717560320, journalId=1296125453100220459, articleId=1297211746784141990, doi=null, pmid=null, pmcid=null, year=null, volume=null, issue=null, pageStart=null, pageEnd=null, url=null, language=null, rfNumber=46, rfOrder=47, authorNames=null, journalName=null, refType=null, unstructuredReference=KE G L, MENG Q, FINLEY T, et al. LightGBM: A highly efficient gradient boosting decision tree[C]// 31st Conference on Neural Information Processing Systems (NIPS). Red Hook, NY, USA: Curran Associates, Inc. , 2017., articleTitle=null, refAbstract=null), Reference(id=1299828229666271613, tenantId=1146029695717560320, journalId=1296125453100220459, articleId=1297211746784141990, doi=null, pmid=null, pmcid=null, year=null, volume=null, issue=null, pageStart=null, pageEnd=null, url=null, language=null, rfNumber=47, rfOrder=48, authorNames=null, journalName=null, refType=null, unstructuredReference=YIN X Q, WU D H, SHANG Y Y, et al. Using an efficientnet-LSTM for the recognition of single cow’s motion behaviours in a complicated environment[J]. Computers and Electronics in Agriculture, 2020, 177: 105707., articleTitle=null, refAbstract=null), Reference(id=1299828229766934910, tenantId=1146029695717560320, journalId=1296125453100220459, articleId=1297211746784141990, doi=null, pmid=null, pmcid=null, year=null, volume=null, issue=null, pageStart=null, pageEnd=null, url=null, language=null, rfNumber=48, rfOrder=49, authorNames=null, journalName=null, refType=null, unstructuredReference=JIANG P, CHEN J J. Displacement prediction of landslide based on generalized regression neural networks with K-fold cross-validation[J]. Neurocomputing, 2016, 198(19): 40-47., articleTitle=null, refAbstract=null), Reference(id=1299828229859209599, tenantId=1146029695717560320, journalId=1296125453100220459, articleId=1297211746784141990, doi=null, pmid=null, pmcid=null, year=null, volume=null, issue=null, pageStart=null, pageEnd=null, url=null, language=null, rfNumber=49, rfOrder=50, authorNames=null, journalName=null, refType=null, unstructuredReference=朱明, 张镇府, 黄凰, 等. 基于轻量级神经网络MobileNetV3-Small的鲈鱼摄食状态分类[J]. 农业工程学报, 2021, 37(19): 165-172., articleTitle=null, refAbstract=null), Reference(id=1299828229934707072, tenantId=1146029695717560320, journalId=1296125453100220459, articleId=1297211746784141990, doi=null, pmid=null, pmcid=null, year=null, volume=null, issue=null, pageStart=null, pageEnd=null, url=null, language=null, rfNumber=49, rfOrder=51, authorNames=null, journalName=null, refType=null, unstructuredReference=ZHU Ming, ZHANG Zhenfu, HUANG Huang, et al. Classification of perch ingesting condition using lightweight neural network MobileNetV3-Small[J]. Transactions of the Chinese Society of Agricultural Engineering (Transactions of the CSAE), 2021, 37(19): 165-172. (in Chinese with English abstract), articleTitle=null, refAbstract=null), Reference(id=1299828230026981761, tenantId=1146029695717560320, journalId=1296125453100220459, articleId=1297211746784141990, doi=null, pmid=null, pmcid=null, year=null, volume=null, issue=null, pageStart=null, pageEnd=null, url=null, language=null, rfNumber=50, rfOrder=52, authorNames=null, journalName=null, refType=null, unstructuredReference=赵思琪, 赵三琴, 国振淇, 等. 池塘水产养殖精准投喂系统研制与试验[J]. 农业工程学报, 2023, 39(22): 27-34., articleTitle=null, refAbstract=null), Reference(id=1299828230115062146, tenantId=1146029695717560320, journalId=1296125453100220459, articleId=1297211746784141990, doi=null, pmid=null, pmcid=null, year=null, volume=null, issue=null, pageStart=null, pageEnd=null, url=null, language=null, rfNumber=50, rfOrder=53, authorNames=null, journalName=null, refType=null, unstructuredReference=ZHAO Siqi, ZHAO Sanqin, GUO Zhenqi, et al. Development of the precise feeding system for pond culture[J]. Transactions of the Chinese Society of Agricultural Engineering (Transactions of the CSAE), 2023, 39(22): 27-34. (in Chinese with English abstract), articleTitle=null, refAbstract=null)], funds=null, companyList=[AuthorCompany(id=1299828221177000188, tenantId=1146029695717560320, journalId=1296125453100220459, articleId=1297211746784141990, xref=1, ext=[AuthorCompanyExt(id=1299828221185388797, tenantId=1146029695717560320, journalId=1296125453100220459, articleId=1297211746784141990, companyId=1299828221177000188, language=EN, country=null, province=null, city=null, postcode=null, companyName=null, departmentName=null, remark=1College of Engineering, Huazhong Agricultural University, Wuhan 430070, China), AuthorCompanyExt(id=1299828221193777406, tenantId=1146029695717560320, journalId=1296125453100220459, articleId=1297211746784141990, companyId=1299828221177000188, language=CN, country=null, province=null, city=null, postcode=null, companyName=null, departmentName=null, remark=1华中农业大学工学院,武汉 430070)]), AuthorCompany(id=1299828221269274879, tenantId=1146029695717560320, journalId=1296125453100220459, articleId=1297211746784141990, xref=2, ext=[AuthorCompanyExt(id=1299828221277663488, tenantId=1146029695717560320, journalId=1296125453100220459, articleId=1297211746784141990, companyId=1299828221269274879, language=EN, country=null, province=null, city=null, postcode=null, companyName=null, departmentName=null, remark=2The Key Laboratory of Aquaculture Facilities Engineering, Ministry of Agriculture and Rural Affairs, Wuhan 430070, China), AuthorCompanyExt(id=1299828221281857793, tenantId=1146029695717560320, journalId=1296125453100220459, articleId=1297211746784141990, companyId=1299828221269274879, language=CN, country=null, province=null, city=null, postcode=null, companyName=null, departmentName=null, remark=2农业农村部水产养殖设施工程重点实验室,武汉 430070)])], figs=[ArticleFig(id=1299828223177683236, tenantId=1146029695717560320, journalId=1296125453100220459, articleId=1297211746784141990, language=EN, label=Fig.1, caption=Schematic diagram of fish feeding audio and video information collection system, figureFileSmall=9WjBPrvl5CTSqpFEt+80ow==, figureFileBig=qgzk7i0N+abfQh89VenT7Q==, tableContent=null), ArticleFig(id=1299828223244792101, tenantId=1146029695717560320, journalId=1296125453100220459, articleId=1297211746784141990, language=CN, label=图1, caption=鱼类摄食音频与视频信息采集系统示意图, figureFileSmall=9WjBPrvl5CTSqpFEt+80ow==, figureFileBig=qgzk7i0N+abfQh89VenT7Q==, tableContent=null), ArticleFig(id=1299828223353844006, tenantId=1146029695717560320, journalId=1296125453100220459, articleId=1297211746784141990, language=EN, label=Fig.2, caption=Flow chart of feeding audio endpoint detection algorithm, figureFileSmall=Iu2fV9ox83H6xcqoZGkuqw==, figureFileBig=UBH9hCeIEUeX+PRUdOFK1g==, tableContent=null), ArticleFig(id=1299828223429341479, tenantId=1146029695717560320, journalId=1296125453100220459, articleId=1297211746784141990, language=CN, label=图2, caption=摄食音频终点检测算法流程图

注:T = 300 为初始阈值,i 表示提取片段的索引,E表示摄食端点。

, figureFileSmall=Iu2fV9ox83H6xcqoZGkuqw==, figureFileBig=UBH9hCeIEUeX+PRUdOFK1g==, tableContent=null), ArticleFig(id=1299828223500644648, tenantId=1146029695717560320, journalId=1296125453100220459, articleId=1297211746784141990, language=EN, label=Fig.3, caption=Framework of a time series-based feeding status prediction model, figureFileSmall=zOLODMPyaedm62wSBr80Fg==, figureFileBig=/846Cg/uF7/DmBKJjOXRyA==, tableContent=null), ArticleFig(id=1299828223567753513, tenantId=1146029695717560320, journalId=1296125453100220459, articleId=1297211746784141990, language=CN, label=图3, caption=基于时间序列的投喂状态预测模型框架, figureFileSmall=zOLODMPyaedm62wSBr80Fg==, figureFileBig=/846Cg/uF7/DmBKJjOXRyA==, tableContent=null), ArticleFig(id=1299828223639056682, tenantId=1146029695717560320, journalId=1296125453100220459, articleId=1297211746784141990, language=EN, label=Fig.4, caption=1D-CNN-based feature extraction network, figureFileSmall=Ndhl5SwKfZr1wIG6CdXJfA==, figureFileBig=QstCQNySHPqVTNmEiylB0w==, tableContent=null), ArticleFig(id=1299828223701971243, tenantId=1146029695717560320, journalId=1296125453100220459, articleId=1297211746784141990, language=CN, label=图4, caption=基于1D-CNN的特征提取网络, figureFileSmall=Ndhl5SwKfZr1wIG6CdXJfA==, figureFileBig=QstCQNySHPqVTNmEiylB0w==, tableContent=null), ArticleFig(id=1299828223756497196, tenantId=1146029695717560320, journalId=1296125453100220459, articleId=1297211746784141990, language=EN, label=Fig.5, caption=LSTM-based feature extraction network, figureFileSmall=09gBckFTlPH6H1P/KMKWZg==, figureFileBig=qhPoWbHCMylHUKjKo+z2GA==, tableContent=null), ArticleFig(id=1299828223815217453, tenantId=1146029695717560320, journalId=1296125453100220459, articleId=1297211746784141990, language=CN, label=图5, caption=基于LSTM的特征提取网络

注:Ct−1Ct分别为上一时间步和当前时间步的记忆细胞,$\tilde C_t $为当前时间步候选记忆细胞,σ为sigmoid 激活函数,tanh为双曲正切激活函数,Ht−1Ht分别为上一时间步和当前时间步的隐藏状态,zt为当前时间步的输入,FtItOt分别为当前时间步遗忘门、输入门和输出门的输出。下同。

, figureFileSmall=09gBckFTlPH6H1P/KMKWZg==, figureFileBig=qhPoWbHCMylHUKjKo+z2GA==, tableContent=null), ArticleFig(id=1299828223890714926, tenantId=1146029695717560320, journalId=1296125453100220459, articleId=1297211746784141990, language=EN, label=Fig.6, caption=GRU-based feature extraction network, figureFileSmall=C6j3GqPW9yQH9JmonqWx1g==, figureFileBig=/42U13jktmuIiPoF8E9Nxg==, tableContent=null), ArticleFig(id=1299828223970406703, tenantId=1146029695717560320, journalId=1296125453100220459, articleId=1297211746784141990, language=CN, label=图6, caption=基于GRU的特征提取网络

注:rt是重置门的输出,ut是更新门的输出,$\hat h_t $为当前时间步的候选隐藏状态。

, figureFileSmall=C6j3GqPW9yQH9JmonqWx1g==, figureFileBig=/42U13jktmuIiPoF8E9Nxg==, tableContent=null), ArticleFig(id=1299828224041709872, tenantId=1146029695717560320, journalId=1296125453100220459, articleId=1297211746784141990, language=EN, label=Fig.7, caption=Transformer-based feature extraction network, figureFileSmall=JNEhT/nA7fyvtUncTMljjA==, figureFileBig=y71IQzdx4Z8OgKaKN0gCVw==, tableContent=null), ArticleFig(id=1299828224108818737, tenantId=1146029695717560320, journalId=1296125453100220459, articleId=1297211746784141990, language=CN, label=图7, caption=基于Transformer的特征提取网络

注:Nx为编码器堆叠的层数。

, figureFileSmall=JNEhT/nA7fyvtUncTMljjA==, figureFileBig=y71IQzdx4Z8OgKaKN0gCVw==, tableContent=null), ArticleFig(id=1299828224192704818, tenantId=1146029695717560320, journalId=1296125453100220459, articleId=1297211746784141990, language=EN, label=Fig.8, caption=Schematic of modelling process, figureFileSmall=v71LGbm+RqkG77dMPpzVKw==, figureFileBig=1/IzfV3aUKeVYLp5kB26mA==, tableContent=null), ArticleFig(id=1299828224276590899, tenantId=1146029695717560320, journalId=1296125453100220459, articleId=1297211746784141990, language=CN, label=图8, caption=建模过程示意图, figureFileSmall=v71LGbm+RqkG77dMPpzVKw==, figureFileBig=1/IzfV3aUKeVYLp5kB26mA==, tableContent=null), ArticleFig(id=1299828224343699764, tenantId=1146029695717560320, journalId=1296125453100220459, articleId=1297211746784141990, language=EN, label=Fig.9, caption=Results of importance ranking of features using two algorithms, figureFileSmall=5bj3931Ko94q5Ix7aLl+uQ==, figureFileBig=uk3pBYwNMZpl9Mod2WSYjg==, tableContent=null), ArticleFig(id=1299828224427585845, tenantId=1146029695717560320, journalId=1296125453100220459, articleId=1297211746784141990, language=CN, label=图9, caption=两种算法的特征重要性排序结果, figureFileSmall=5bj3931Ko94q5Ix7aLl+uQ==, figureFileBig=uk3pBYwNMZpl9Mod2WSYjg==, tableContent=null), ArticleFig(id=1299828224503083318, tenantId=1146029695717560320, journalId=1296125453100220459, articleId=1297211746784141990, language=EN, label=Fig.10, caption=Confusion matrix of the test results of feeding status using two features, figureFileSmall=3j+EOO6+LIVTKsVxml6dKA==, figureFileBig=q3brf2esFKN49chU6JHNFg==, tableContent=null), ArticleFig(id=1299828224591163703, tenantId=1146029695717560320, journalId=1296125453100220459, articleId=1297211746784141990, language=CN, label=图10, caption=使用两种特征投喂状态的测试结果混淆矩阵, figureFileSmall=3j+EOO6+LIVTKsVxml6dKA==, figureFileBig=q3brf2esFKN49chU6JHNFg==, tableContent=null), ArticleFig(id=1299828224670855480, tenantId=1146029695717560320, journalId=1296125453100220459, articleId=1297211746784141990, language=EN, label=Fig.11, caption=Pond-based captive system, figureFileSmall=fPHOmbKi9DIaKMt3+wJb5Q==, figureFileBig=VYgIGWQJAA9524mS7GXzIw==, tableContent=null), ArticleFig(id=1299828224733770041, tenantId=1146029695717560320, journalId=1296125453100220459, articleId=1297211746784141990, language=CN, label=图11, caption=池塘圈养系统, figureFileSmall=fPHOmbKi9DIaKMt3+wJb5Q==, figureFileBig=VYgIGWQJAA9524mS7GXzIw==, tableContent=null), ArticleFig(id=1299828224809267514, tenantId=1146029695717560320, journalId=1296125453100220459, articleId=1297211746784141990, language=EN, label=Fig.12, caption=Flow chart of the feeding decision, figureFileSmall=UoShcceDa01ni7ccwnyLXw==, figureFileBig=yaEw33KCPSpBNdv0xXSZDA==, tableContent=null), ArticleFig(id=1299828224880570683, tenantId=1146029695717560320, journalId=1296125453100220459, articleId=1297211746784141990, language=CN, label=图12, caption=投喂决策程序流程图

注:M0为理论投喂量,M为累计投喂量,m为单轮投喂量。

, figureFileSmall=UoShcceDa01ni7ccwnyLXw==, figureFileBig=yaEw33KCPSpBNdv0xXSZDA==, tableContent=null), ArticleFig(id=1299828224947679548, tenantId=1146029695717560320, journalId=1296125453100220459, articleId=1297211746784141990, language=EN, label=Tab.1, caption=

Grading of fish feeding intensity

, figureFileSmall=null, figureFileBig=null, tableContent=
摄食行为水平
Level of feeding
behavior
鱼类摄食行为
Fish feeding behavior
数据集类别
Dataset category
强Strong鱼群主动摄食而且运动范围大CF
中Medium鱼群开始主动摄食但是运动范围小
弱Weak鱼群只对附近饵料有反应SF
无None鱼群对饵料无反应
), ArticleFig(id=1299828225039954237, tenantId=1146029695717560320, journalId=1296125453100220459, articleId=1297211746784141990, language=CN, label=表1, caption=

鱼类摄食强度分级

, figureFileSmall=null, figureFileBig=null, tableContent=
摄食行为水平
Level of feeding
behavior
鱼类摄食行为
Fish feeding behavior
数据集类别
Dataset category
强Strong鱼群主动摄食而且运动范围大CF
中Medium鱼群开始主动摄食但是运动范围小
弱Weak鱼群只对附近饵料有反应SF
无None鱼群对饵料无反应
), ArticleFig(id=1299828225111257406, tenantId=1146029695717560320, journalId=1296125453100220459, articleId=1297211746784141990, language=EN, label=Tab.2, caption=

Correlation analysis between feeding features

, figureFileSmall=null, figureFileBig=null, tableContent=
特征
Feature
参数
Parameter
ZCRMSE
注:r 为相关系数;P为相关系数的显著性概率水平;***表示在0.001水平上显著相关。
Note: r is the correlation coefficient and P is the significance probability level of the correlation coefficient; ***indicate significant at 0.001 level.
PWr−0.653***0.784***
P<0.001<0.001
ZCRr−0.759***
P<0.001
), ArticleFig(id=1299828225178366271, tenantId=1146029695717560320, journalId=1296125453100220459, articleId=1297211746784141990, language=CN, label=表2, caption=

摄食特征相关性分析

, figureFileSmall=null, figureFileBig=null, tableContent=
特征
Feature
参数
Parameter
ZCRMSE
注:r 为相关系数;P为相关系数的显著性概率水平;***表示在0.001水平上显著相关。
Note: r is the correlation coefficient and P is the significance probability level of the correlation coefficient; ***indicate significant at 0.001 level.
PWr−0.653***0.784***
P<0.001<0.001
ZCRr−0.759***
P<0.001
), ArticleFig(id=1299828225249669440, tenantId=1146029695717560320, journalId=1296125453100220459, articleId=1297211746784141990, language=EN, label=Tab.3, caption=

Principal component analysis results

, figureFileSmall=null, figureFileBig=null, tableContent=
主成分
Number of
components
特征值
Eigenvalue
累计方差贡献率
Cumulative variance
contribution/%
成分系数矩阵
Component coefficient matrix
X1X2X3
注:X1X3为PW、ZCR、MSE的标准分。
Note: X1-X3 are PW, ZCR, MSE standard scores.
12.46082.1660.363−0.3590.380
20.34993.7891.1251.2610.115
30.191100.000−1.1190.8291.851
), ArticleFig(id=1299828225316778305, tenantId=1146029695717560320, journalId=1296125453100220459, articleId=1297211746784141990, language=CN, label=表3, caption=

主成分分析结果

, figureFileSmall=null, figureFileBig=null, tableContent=
主成分
Number of
components
特征值
Eigenvalue
累计方差贡献率
Cumulative variance
contribution/%
成分系数矩阵
Component coefficient matrix
X1X2X3
注:X1X3为PW、ZCR、MSE的标准分。
Note: X1-X3 are PW, ZCR, MSE standard scores.
12.46082.1660.363−0.3590.380
20.34993.7891.1251.2610.115
30.191100.000−1.1190.8291.851
), ArticleFig(id=1299828225379692866, tenantId=1146029695717560320, journalId=1296125453100220459, articleId=1297211746784141990, language=EN, label=Tab.4, caption=

Classification performance of models with different input features for Micropterus salmoides feeding status classification

, figureFileSmall=null, figureFileBig=null, tableContent=
模型
Model
输入特征
Input features
投喂状态
Feeding status
准确率
Accuracy
精确率
Precision
召回率
Recall
F1分数
F1
GBDT原始特征
Original features
CF0.96170.960.980.97
SF0.960.910.93
Mean0.96170.960.9450.95
敏感特征
Sensitive features
CF0.96930.980.980.98
SF0.950.950.95
Mean0.96930.9650.9650.97
RF原始特征
Original features
CF0.95790.970.970.97
SF0.930.920.93
Mean0.95790.950.9450.95
敏感特征
Sensitive features
CF0.96170.980.970.97
SF0.930.950.94
Mean0.96170.960.960.96
), ArticleFig(id=1299828225446801731, tenantId=1146029695717560320, journalId=1296125453100220459, articleId=1297211746784141990, language=CN, label=表4, caption=

大口黑鲈投喂状态分类中不同输入特征的模型分类性能比较

, figureFileSmall=null, figureFileBig=null, tableContent=
模型
Model
输入特征
Input features
投喂状态
Feeding status
准确率
Accuracy
精确率
Precision
召回率
Recall
F1分数
F1
GBDT原始特征
Original features
CF0.96170.960.980.97
SF0.960.910.93
Mean0.96170.960.9450.95
敏感特征
Sensitive features
CF0.96930.980.980.98
SF0.950.950.95
Mean0.96930.9650.9650.97
RF原始特征
Original features
CF0.95790.970.970.97
SF0.930.920.93
Mean0.95790.950.9450.95
敏感特征
Sensitive features
CF0.96170.980.970.97
SF0.930.950.94
Mean0.96170.960.960.96
), ArticleFig(id=1299828225534882116, tenantId=1146029695717560320, journalId=1296125453100220459, articleId=1297211746784141990, language=EN, label=Tab.5, caption=

Impact of time window selection on feeding status classification models

, figureFileSmall=null, figureFileBig=null, tableContent=
模型
Model
时间窗口
Time
window
投喂状态
Feeding
status
准确率
Accuracy
精确率
Precision
召回率
Recall
F1分数
F1
1D-CNN-RC1CF0.930.940.970.95
SF0.930.860.89
Mean0.930.9350.9150.92
2CF0.920.960.910.94
SF0.850.940.89
Mean0.920.9050.9250.915
3CF0.930.940.940.94
SF0.900.910.91
Mean0.930.920.9250.925
LSTM-RC1CF0.930.950.950.95
SF0.900.900.90
Mean0.930.9250.9250.925
2CF0.920.960.930.94
SF0.870.920.89
Mean0.920.9150.9250.915
3CF0.930.940.940.94
SF0.900.910.91
Mean0.930.920.9250.925
GRU-RC1CF0.930.960.940.95
SF0.880.910.89
Mean0.930.920.9250.92
2CF0.920.970.910.94
SF0.850.950.90
Mean0.920.910.930.92
3CF0.930.960.930.94
SF0.900.910.91
Mean0.930.930.920.925
Transformer- RC1CF0.940.950.960.96
SF0.920.900.91
Mean0.940.9350.930.935
2CF0.930.970.930.95
SF0.870.950.91
Mean0.930.920.940.93
3CF0.940960.940.95
SF0.920.940.93
Mean0.940.940.940.94
), ArticleFig(id=1299828225606185285, tenantId=1146029695717560320, journalId=1296125453100220459, articleId=1297211746784141990, language=CN, label=表5, caption=

时间窗口选择对投喂状态分类模型的影响

, figureFileSmall=null, figureFileBig=null, tableContent=
模型
Model
时间窗口
Time
window
投喂状态
Feeding
status
准确率
Accuracy
精确率
Precision
召回率
Recall
F1分数
F1
1D-CNN-RC1CF0.930.940.970.95
SF0.930.860.89
Mean0.930.9350.9150.92
2CF0.920.960.910.94
SF0.850.940.89
Mean0.920.9050.9250.915
3CF0.930.940.940.94
SF0.900.910.91
Mean0.930.920.9250.925
LSTM-RC1CF0.930.950.950.95
SF0.900.900.90
Mean0.930.9250.9250.925
2CF0.920.960.930.94
SF0.870.920.89
Mean0.920.9150.9250.915
3CF0.930.940.940.94
SF0.900.910.91
Mean0.930.920.9250.925
GRU-RC1CF0.930.960.940.95
SF0.880.910.89
Mean0.930.920.9250.92
2CF0.920.970.910.94
SF0.850.950.90
Mean0.920.910.930.92
3CF0.930.960.930.94
SF0.900.910.91
Mean0.930.930.920.925
Transformer- RC1CF0.940.950.960.96
SF0.920.900.91
Mean0.940.9350.930.935
2CF0.930.970.930.95
SF0.870.950.91
Mean0.930.920.940.93
3CF0.940960.940.95
SF0.920.940.93
Mean0.940.940.940.94
), ArticleFig(id=1299828225664905542, tenantId=1146029695717560320, journalId=1296125453100220459, articleId=1297211746784141990, language=EN, label=Tab.6, caption=

Performance of 8 ML classification models

, figureFileSmall=null, figureFileBig=null, tableContent=
模型
Model
投喂状态
Feeding status
准确率
Accuracy
精确率
Precision
召回率
Recall
F1分数
F1
KNNCF0.930.960.940.95
SF0.860.900.88
Mean0.930.920.920.915
DTCF0.930.950.950.95
SF0.880.880.88
Mean0.930.920.920.92
SVMCF0.950.960.970.96
SF0.920.910.92
Mean0.950.940.940.94
RFCF0.960.980.970.97
SF0.930.950.94
Mean0.960.9550.960.955
AdaBoostCF0.970.980.970.98
SF0.940.960.95
Mean0.970.960.9650.965
GBDTCF0.970.980.980.98
SF0.950.950.95
Mean0.970.9650.9650.965
XGBoostCF0.970.990.970.98
SF0.940.970.96
Mean0.970.9650.970.97
LightGBMCF0.980.990.970.98
SF0.940.990.96
Mean0.980.9650.980.97
), ArticleFig(id=1299828225744597319, tenantId=1146029695717560320, journalId=1296125453100220459, articleId=1297211746784141990, language=CN, label=表6, caption=

八种机器学习分类模型的性能

, figureFileSmall=null, figureFileBig=null, tableContent=
模型
Model
投喂状态
Feeding status
准确率
Accuracy
精确率
Precision
召回率
Recall
F1分数
F1
KNNCF0.930.960.940.95
SF0.860.900.88
Mean0.930.920.920.915
DTCF0.930.950.950.95
SF0.880.880.88
Mean0.930.920.920.92
SVMCF0.950.960.970.96
SF0.920.910.92
Mean0.950.940.940.94
RFCF0.960.980.970.97
SF0.930.950.94
Mean0.960.9550.960.955
AdaBoostCF0.970.980.970.98
SF0.940.960.95
Mean0.970.960.9650.965
GBDTCF0.970.980.980.98
SF0.950.950.95
Mean0.970.9650.9650.965
XGBoostCF0.970.990.970.98
SF0.940.970.96
Mean0.970.9650.970.97
LightGBMCF0.980.990.970.98
SF0.940.990.96
Mean0.980.9650.980.97
), ArticleFig(id=1299828225815900488, tenantId=1146029695717560320, journalId=1296125453100220459, articleId=1297211746784141990, language=EN, label=Tab.7, caption=

Results of the model in the validation experiment

, figureFileSmall=null, figureFileBig=null, tableContent=
模型
Model
时间
窗口
Time
window
投喂状态
Feeding
status
准确率
Accuracy
精确率Precision召回率RecallF1分数
F1
Transformer- RC1CF0.920.910.970.94
SF0.940.810.87
Mean0.920.9250.890.91
3CF0.900.920.900.91
SF0.880.900.89
Mean0.900.900.900.90
LightGBM-CF0.950.980.940.96
SF0.880.960.92
Mean0.950.930.950.94
), ArticleFig(id=1299828225887203657, tenantId=1146029695717560320, journalId=1296125453100220459, articleId=1297211746784141990, language=CN, label=表7, caption=

模型在验证试验中的结果

, figureFileSmall=null, figureFileBig=null, tableContent=
模型
Model
时间
窗口
Time
window
投喂状态
Feeding
status
准确率
Accuracy
精确率Precision召回率RecallF1分数
F1
Transformer- RC1CF0.920.910.970.94
SF0.940.810.87
Mean0.920.9250.890.91
3CF0.900.920.900.91
SF0.880.900.89
Mean0.900.900.900.90
LightGBM-CF0.950.980.940.96
SF0.880.960.92
Mean0.950.930.950.94
), ArticleFig(id=1299828225954312522, tenantId=1146029695717560320, journalId=1296125453100220459, articleId=1297211746784141990, language=EN, label=Tab.8, caption=

Performance of fish growth for two feeding model

, figureFileSmall=null, figureFileBig=null, tableContent=
规格
Size
模式
Mode
数量
Number/尾
M1/kgM2/kgF/kgFCRWGR/%SGR/%
注:饵料系数(feed conversion ratio, FCR)、增重率(weight gain ratio, WGR)和特定生长率(specific growth rate, SGR),F为试验期间的总投喂量,kg;M1M2分别表示投喂试验开始和结束时大口黑鲈鱼群的总质量,kg。
Note: F is the total feed amount during the experiment, kg; M1 and M2 represent the total mass of the largemouth bass population at the beginning and of the feeding trial, kg, respectively.
小规格鱼
Small-size
fish
对照组1808.169.823.922.3720.291.32
试验组10.5112.262.431.3916.671.10
大规格鱼
Large-size
fish
对照组6316.5817.832.992.407.500.52
试验组17.3318.992.381.439.610.66
), ArticleFig(id=1299828226017227083, tenantId=1146029695717560320, journalId=1296125453100220459, articleId=1297211746784141990, language=CN, label=表8, caption=

两种投喂模式下大口黑鲈的生长参数

, figureFileSmall=null, figureFileBig=null, tableContent=
规格
Size
模式
Mode
数量
Number/尾
M1/kgM2/kgF/kgFCRWGR/%SGR/%
注:饵料系数(feed conversion ratio, FCR)、增重率(weight gain ratio, WGR)和特定生长率(specific growth rate, SGR),F为试验期间的总投喂量,kg;M1M2分别表示投喂试验开始和结束时大口黑鲈鱼群的总质量,kg。
Note: F is the total feed amount during the experiment, kg; M1 and M2 represent the total mass of the largemouth bass population at the beginning and of the feeding trial, kg, respectively.
小规格鱼
Small-size
fish
对照组1808.169.823.922.3720.291.32
试验组10.5112.262.431.3916.671.10
大规格鱼
Large-size
fish
对照组6316.5817.832.992.407.500.52
试验组17.3318.992.381.439.610.66
)], attaches=null, journal=Journal(id=1296123985890738214, delFlag=0, nameCn=农业工程学报, nameEn=Transactions of the Chinese Society of Agricultural Engineering, nameHistory1=null, nameHistory2=null, issn=1002-6819, eissn=null, cn=11-2047/S, coden=null, periodic=3, language=CN, oaType=null, ccby=null, superviseOffice=null, ownerOffice=null, pubOffice=null, editorOffice=null, officeType=null, aims=null, clcCode=null, officeProv=null, officeCity=null, officeAddr=null, officeZip=null, officeEmail=null, officePhone=null, editDirector=null, officeDirector=null, officeDirectorPhone=null, officeStaffNum=null, officeEmpNum=null, coverPicUrl=+90GfMrDOo0sGnhM27IfIQ==, journalPrice=null, startedYear=null, abbrevIsoEn=Transactions of the Chinese Society of Agricultural Engineering, journalRemark=null, publicationField=null, createdTime=1786949639057, updatedTime=1788949514636, createdBy=18614031015, updatedBy=13041195026, firstLetterCn=N, firstLetterEn=N, subjectCode=Agriculture, subjectName=null, subjectCodeEn=Agriculture, subjectNameEn=null, picCn=+90GfMrDOo0sGnhM27IfIQ==, picEn=+UrL7Qgb+cjoX8OpJtziCA==, jcr=null, cjcr=null, exts=[JournalExt(id=1304512072117678568, language=CN, name=农业工程学报, nameHistory1=null, nameHistory2=null, managedBy=, sponsoredBy=, publishedBy=, editorOffice=, officeProv=null, officeCity=null, officeAddr=, officeZip=, editDirector=, officeDirector=null, officePhone=null, coverPicUrl=null, journalRemark=, submitArticleUrl=null, websiteUrl=, createdTime=1788949514655, updatedTime=1788949514655, createdBy=13041195026, updatedBy=13041195026, submissionGuidelinesUrl=, submissionAuthorUrl=http://tsg.tcsae.org/nygcxb/author/login, submissionEditorUrl=https://nygcxb.ijournals.cn/nygcxb/editor/login, submissionReviewUrl=http://tsg.tcsae.org/nygcxb/reviewer/login, submissionCeEditorUrl=, submissionAeEditorUrl=, option={"copyright":""}), JournalExt(id=1304512072168010217, language=EN, name=Transactions of the Chinese Society of Agricultural Engineering, nameHistory1=null, nameHistory2=null, managedBy=, sponsoredBy=, publishedBy=, editorOffice=, officeProv=null, officeCity=null, officeAddr=, officeZip=, editDirector=, officeDirector=null, officePhone=null, coverPicUrl=null, journalRemark=, submitArticleUrl=null, websiteUrl=, createdTime=1788949514667, updatedTime=1788949514667, createdBy=13041195026, updatedBy=13041195026, submissionGuidelinesUrl=, submissionAuthorUrl=http://tsg.tcsae.org/nygcxb/author/login, submissionEditorUrl=https://nygcxb.ijournals.cn/nygcxb/editor/login, submissionReviewUrl=http://tsg.tcsae.org/nygcxb/reviewer/login, submissionCeEditorUrl=, submissionAeEditorUrl=, option={"copyright":""})], databaseList=null, tenantJournalId=1296125453100220459, websiteList=[Website(id=1296128922534564555, webName=null, webTitle=null, webDomain=null, webCopyrigh=null, webIpcNo=null, seoTitle=null, seoKeywords=null, seoDescription=null, tenantJournalId=null, journalId=1296125453100220459, journalNameCn=null, journalNameEn=null, grayFlag=null, tenantId=1146029695717560320, platformId=null, journalGroupId=null, journalGroupNameCn=null, journalGroupNameEn=null, type=1, domain=https://castjournals.cast.org.cn/joweb/nygcxb/CN, language=CN, createTime=1786950816043, createBy=18614031015, updateTime=1786950989293, updateBy=18614031015, name=农业工程学报-中文, tplId=1146099689490845704, title=农业工程学报, delFlag=0, indexPage=/home, props=[WebsiteProps(id=1296129761215009609, tenantId=1146029695717560320, journalId=null, journalGroupId=null, siteId=1296128922534564555, code=articleTextType, value=kx, createTime=1786951016000, updateTime=1786951016000, creator=18614031015, updator=18614031015), WebsiteProps(id=1296129761194038086, tenantId=1146029695717560320, journalId=null, journalGroupId=null, siteId=1296128922534564555, code=banner, value=null, createTime=1786951015995, updateTime=1786951015995, creator=18614031015, updator=18614031015), WebsiteProps(id=1296129761231786828, tenantId=1146029695717560320, journalId=null, journalGroupId=null, siteId=1296128922534564555, code=grayFlag, value=0, createTime=1786951016004, updateTime=1786951016004, creator=18614031015, updator=18614031015), WebsiteProps(id=1296129761185649477, tenantId=1146029695717560320, journalId=null, journalGroupId=null, siteId=1296128922534564555, code=logo, value=https://castjournals.cast.org.cn/joweb/nygcxb/CN/file/pic?fileId=6wW+hIE4oBGmmFvGgEb5JQ==, createTime=1786951015993, updateTime=1786951015993, creator=18614031015, updator=18614031015), WebsiteProps(id=1296129761244369742, tenantId=1146029695717560320, journalId=null, journalGroupId=null, siteId=1296128922534564555, code=minRunFlag, value=0, createTime=1786951016007, updateTime=1786951016007, creator=18614031015, updator=18614031015), WebsiteProps(id=1296129761206621000, tenantId=1146029695717560320, journalId=null, journalGroupId=null, siteId=1296128922534564555, code=picServerUrl, value=https://castjournals.cast.org.cn/joweb/nygcxb/CN/file/pic, createTime=1786951015998, updateTime=1786951015998, creator=18614031015, updator=18614031015), WebsiteProps(id=1296129761240175437, tenantId=1146029695717560320, journalId=null, journalGroupId=null, siteId=1296128922534564555, code=silenceFlag, value=0, createTime=1786951016006, updateTime=1786951016006, creator=18614031015, updator=18614031015), WebsiteProps(id=1296129761202426695, tenantId=1146029695717560320, journalId=null, journalGroupId=null, siteId=1296128922534564555, code=staticResourcePath, value=https://castjournals.cast.org.cn/joweb/cast_kjdb_cn_619/, createTime=1786951015997, updateTime=1786951015997, creator=18614031015, updator=18614031015), WebsiteProps(id=1296129761219203914, tenantId=1146029695717560320, journalId=null, journalGroupId=null, siteId=1296128922534564555, code=themeColor, value=null, createTime=1786951016001, updateTime=1786951016001, creator=18614031015, updator=18614031015), WebsiteProps(id=1296129761227592523, tenantId=1146029695717560320, journalId=null, journalGroupId=null, siteId=1296128922534564555, code=themeStyle, value=null, createTime=1786951016003, updateTime=1786951016003, creator=18614031015, updator=18614031015)]), Website(id=1296128922677170905, webName=null, webTitle=null, webDomain=null, webCopyrigh=null, webIpcNo=null, seoTitle=null, seoKeywords=null, seoDescription=null, tenantJournalId=null, journalId=1296125453100220459, journalNameCn=null, journalNameEn=null, grayFlag=null, tenantId=1146029695717560320, platformId=null, journalGroupId=null, journalGroupNameCn=null, journalGroupNameEn=null, type=1, domain=https://castjournals.cast.org.cn/joweb/nygcxb/EN, language=EN, createTime=1786950816077, createBy=18614031015, updateTime=1786950985772, updateBy=18614031015, name=农业工程学报-英文, tplId=1146101810881728533, title=Transactions of the Chinese Society of Agricultural Engineering, delFlag=0, indexPage=/home, props=[WebsiteProps(id=1296129731049575225, tenantId=1146029695717560320, journalId=null, journalGroupId=null, siteId=1296128922677170905, code=articleTextType, value=kx, createTime=1786951008808, updateTime=1786951008808, creator=18614031015, updator=18614031015), WebsiteProps(id=1296129731028603702, tenantId=1146029695717560320, journalId=null, journalGroupId=null, siteId=1296128922677170905, code=banner, value=null, createTime=1786951008803, updateTime=1786951008803, creator=18614031015, updator=18614031015), WebsiteProps(id=1296129731070546748, tenantId=1146029695717560320, journalId=null, journalGroupId=null, siteId=1296128922677170905, code=grayFlag, value=0, createTime=1786951008813, updateTime=1786951008813, creator=18614031015, updator=18614031015), WebsiteProps(id=1296129731016020789, tenantId=1146029695717560320, journalId=null, journalGroupId=null, siteId=1296128922677170905, code=logo, value=https://castjournals.cast.org.cn/joweb/nygcxb/EN/file/pic?fileId=6wW+hIE4oBGmmFvGgEb5JQ==, createTime=1786951008800, updateTime=1786951008800, creator=18614031015, updator=18614031015), WebsiteProps(id=1296129731087323966, tenantId=1146029695717560320, journalId=null, journalGroupId=null, siteId=1296128922677170905, code=minRunFlag, value=0, createTime=1786951008817, updateTime=1786951008817, creator=18614031015, updator=18614031015), WebsiteProps(id=1296129731041186616, tenantId=1146029695717560320, journalId=null, journalGroupId=null, siteId=1296128922677170905, code=picServerUrl, value=https://castjournals.cast.org.cn/joweb/nygcxb/EN/file/pic, createTime=1786951008806, updateTime=1786951008806, creator=18614031015, updator=18614031015), WebsiteProps(id=1296129731078935357, tenantId=1146029695717560320, journalId=null, journalGroupId=null, siteId=1296128922677170905, code=silenceFlag, value=0, createTime=1786951008815, updateTime=1786951008815, creator=18614031015, updator=18614031015), WebsiteProps(id=1296129731032798007, tenantId=1146029695717560320, journalId=null, journalGroupId=null, siteId=1296128922677170905, code=staticResourcePath, value=https://castjournals.cast.org.cn/joweb/cast_kjdb_en_623/, createTime=1786951008804, updateTime=1786951008804, creator=18614031015, updator=18614031015), WebsiteProps(id=1296129731053769530, tenantId=1146029695717560320, journalId=null, journalGroupId=null, siteId=1296128922677170905, code=themeColor, value=null, createTime=1786951008809, updateTime=1786951008809, creator=18614031015, updator=18614031015), WebsiteProps(id=1296129731062158139, tenantId=1146029695717560320, journalId=null, journalGroupId=null, siteId=1296128922677170905, code=themeStyle, value=null, createTime=1786951008811, updateTime=1786951008811, creator=18614031015, updator=18614031015)])], journalTitle=农业工程学报, weixinUrl=null, journalUrl=http://www.tcsae.org/, iacademicId=null, status=1, seqNo=null, journalTitleEn=Transactions of the Chinese Society of Agricultural Engineering, journalPhotoCn=+90GfMrDOo0sGnhM27IfIQ==, journalPhotoEn=+UrL7Qgb+cjoX8OpJtziCA==, journalFirstLetter=N, journalRecommend=null, journalNew=null, journalCollection=null, jcrJf=null, cjcrJf=null, jcrJfStr=null, cjcrJfStr=null, submissionFirstDecision=null, sciSubjectClassification=null, casSubjectClassification=null, citeScore=null, totalCitationFrequency=null, icpCode=null, psCode=null, advertisingLicenseCode=null, copyrightInformation=null, country=null, option=, provinceCode=null, provinceName=null, collectFlag=false, interPubPlatform=, interPubPlatformUrl=null), detailUrlCn=https://castjournals.cast.org.cn/joweb/nygcxb/CN/10.11975/j.issn.1002-6819.202601127, detailUrlEn=https://castjournals.cast.org.cn/joweb/nygcxb/EN/10.11975/j.issn.1002-6819.202601127, pdfUrlCn=https://castjournals.cast.org.cn/joweb/nygcxb/CN/PDF/10.11975/j.issn.1002-6819.202601127, pdfUrlEn=https://castjournals.cast.org.cn/joweb/nygcxb/EN/PDF/10.11975/j.issn.1002-6819.202601127, aliStartDate=0, aliEndDate=0, collectionFlag=false, citedCount=null, citedUrl=null, previewStatus=0, delFlag=0, hasFullText=1, orderTime=1782748800000, fullTextJson=null, articleText=null, reference=null)
收藏切换
基于双模型渐进式的大口黑鲈投喂策略
收藏切换
PDF下载
李亚苹 , 谭鹤群 * , 陈以钒 , 张义仁
农业工程学报 | 农产品加工工程 2026,42(12): 384-396
收起
收藏切换
农业工程学报 |农产品加工工程 2026 , 42 (12) : 384 -396
基于双模型渐进式的大口黑鲈投喂策略
全屏
[Author(id=1299828221348966659, tenantId=1146029695717560320, journalId=1296125453100220459, articleId=1297211746784141990, orderNo=0, firstName=null, middleName=null, lastName=null, nameCn=null, orcid=null, stid=null, country=null, authorPic=null, dead=0, email=liyaping@webmail.hzau.edu.cn, emailSecond=null, emailThird=null, correspondingAuthor=0, authorType=1, ext={EN=AuthorExt(id=1299828221424464134, tenantId=1146029695717560320, journalId=1296125453100220459, articleId=1297211746784141990, authorId=1299828221348966659, language=EN, stringName=Yaping LI, firstName=Yaping, middleName=null, lastName=LI, prefix=null, suffix=null, authorComment=null, nameInitials=null, affiliation=null, department=null, xref=null, address=1College of Engineering, Huazhong Agricultural University, Wuhan 430070, China
2The Key Laboratory of Aquaculture Facilities Engineering, Ministry of Agriculture and Rural Affairs, Wuhan 430070, China, bio=null, bioImg=null, bioContent=null, aboutCorrespAuthor=null), CN=AuthorExt(id=1299828221487378695, tenantId=1146029695717560320, journalId=1296125453100220459, articleId=1297211746784141990, authorId=1299828221348966659, language=CN, stringName=李亚苹, firstName=null, middleName=null, lastName=null, prefix=null, suffix=null, authorComment=null, nameInitials=null, affiliation=null, department=null, xref=null, address=1华中农业大学工学院,武汉 430070
2农业农村部水产养殖设施工程重点实验室,武汉 430070, bio={"content":"

李亚苹,博士生,研究方向为现代养殖技术与装备。Email:

"}, bioImg=null, bioContent=

李亚苹,博士生,研究方向为现代养殖技术与装备。Email:

, aboutCorrespAuthor=null)}, companyList=[AuthorCompany(id=1299828221177000188, tenantId=1146029695717560320, journalId=1296125453100220459, articleId=1297211746784141990, xref=1, ext=[AuthorCompanyExt(id=1299828221185388797, tenantId=1146029695717560320, journalId=1296125453100220459, articleId=1297211746784141990, companyId=1299828221177000188, language=EN, country=null, province=null, city=null, postcode=null, companyName=null, departmentName=null, remark=1College of Engineering, Huazhong Agricultural University, Wuhan 430070, China), AuthorCompanyExt(id=1299828221193777406, tenantId=1146029695717560320, journalId=1296125453100220459, articleId=1297211746784141990, companyId=1299828221177000188, language=CN, country=null, province=null, city=null, postcode=null, companyName=null, departmentName=null, remark=1华中农业大学工学院,武汉 430070)]), AuthorCompany(id=1299828221269274879, tenantId=1146029695717560320, journalId=1296125453100220459, articleId=1297211746784141990, xref=2, ext=[AuthorCompanyExt(id=1299828221277663488, tenantId=1146029695717560320, journalId=1296125453100220459, articleId=1297211746784141990, companyId=1299828221269274879, language=EN, country=null, province=null, city=null, postcode=null, companyName=null, departmentName=null, remark=2The Key Laboratory of Aquaculture Facilities Engineering, Ministry of Agriculture and Rural Affairs, Wuhan 430070, China), AuthorCompanyExt(id=1299828221281857793, tenantId=1146029695717560320, journalId=1296125453100220459, articleId=1297211746784141990, companyId=1299828221269274879, language=CN, country=null, province=null, city=null, postcode=null, companyName=null, departmentName=null, remark=2农业农村部水产养殖设施工程重点实验室,武汉 430070)])]), Author(id=1299828221579653385, tenantId=1146029695717560320, journalId=1296125453100220459, articleId=1297211746784141990, orderNo=1, firstName=null, middleName=null, lastName=null, nameCn=null, orcid=null, stid=null, country=null, authorPic=null, dead=0, email=thq@mail.hzau.edu.cn, emailSecond=null, emailThird=null, correspondingAuthor=1, authorType=1, ext={EN=AuthorExt(id=1299828221659345164, tenantId=1146029695717560320, journalId=1296125453100220459, articleId=1297211746784141990, authorId=1299828221579653385, language=EN, stringName=Hequn TAN, firstName=Hequn, middleName=null, lastName=TAN, prefix=null, suffix=null, authorComment=null, nameInitials=null, affiliation=null, department=null, xref=*, address=1College of Engineering, Huazhong Agricultural University, Wuhan 430070, China
2The Key Laboratory of Aquaculture Facilities Engineering, Ministry of Agriculture and Rural Affairs, Wuhan 430070, China, bio=null, bioImg=null, bioContent=null, aboutCorrespAuthor=null), CN=AuthorExt(id=1299828221734842637, tenantId=1146029695717560320, journalId=1296125453100220459, articleId=1297211746784141990, authorId=1299828221579653385, language=CN, stringName=谭鹤群, firstName=null, middleName=null, lastName=null, prefix=null, suffix=null, authorComment=null, nameInitials=null, affiliation=null, department=null, xref=*, address=1华中农业大学工学院,武汉 430070
2农业农村部水产养殖设施工程重点实验室,武汉 430070, bio=null, bioImg=null, bioContent=null, aboutCorrespAuthor=null)}, companyList=[AuthorCompany(id=1299828221177000188, tenantId=1146029695717560320, journalId=1296125453100220459, articleId=1297211746784141990, xref=1, ext=[AuthorCompanyExt(id=1299828221185388797, tenantId=1146029695717560320, journalId=1296125453100220459, articleId=1297211746784141990, companyId=1299828221177000188, language=EN, country=null, province=null, city=null, postcode=null, companyName=null, departmentName=null, remark=1College of Engineering, Huazhong Agricultural University, Wuhan 430070, China), AuthorCompanyExt(id=1299828221193777406, tenantId=1146029695717560320, journalId=1296125453100220459, articleId=1297211746784141990, companyId=1299828221177000188, language=CN, country=null, province=null, city=null, postcode=null, companyName=null, departmentName=null, remark=1华中农业大学工学院,武汉 430070)]), AuthorCompany(id=1299828221269274879, tenantId=1146029695717560320, journalId=1296125453100220459, articleId=1297211746784141990, xref=2, ext=[AuthorCompanyExt(id=1299828221277663488, tenantId=1146029695717560320, journalId=1296125453100220459, articleId=1297211746784141990, companyId=1299828221269274879, language=EN, country=null, province=null, city=null, postcode=null, companyName=null, departmentName=null, remark=2The Key Laboratory of Aquaculture Facilities Engineering, Ministry of Agriculture and Rural Affairs, Wuhan 430070, China), AuthorCompanyExt(id=1299828221281857793, tenantId=1146029695717560320, journalId=1296125453100220459, articleId=1297211746784141990, companyId=1299828221269274879, language=CN, country=null, province=null, city=null, postcode=null, companyName=null, departmentName=null, remark=2农业农村部水产养殖设施工程重点实验室,武汉 430070)])]), Author(id=1299828221814534415, tenantId=1146029695717560320, journalId=1296125453100220459, articleId=1297211746784141990, orderNo=2, firstName=null, middleName=null, lastName=null, nameCn=null, orcid=null, stid=null, country=null, authorPic=null, dead=0, email=null, emailSecond=null, emailThird=null, correspondingAuthor=0, authorType=1, ext={EN=AuthorExt(id=1299828221902614802, tenantId=1146029695717560320, journalId=1296125453100220459, articleId=1297211746784141990, authorId=1299828221814534415, language=EN, stringName=Yifan CHEN, firstName=Yifan, middleName=null, lastName=CHEN, prefix=null, suffix=null, authorComment=null, nameInitials=null, affiliation=null, department=null, xref=null, address=1College of Engineering, Huazhong Agricultural University, Wuhan 430070, China
2The Key Laboratory of Aquaculture Facilities Engineering, Ministry of Agriculture and Rural Affairs, Wuhan 430070, China, bio=null, bioImg=null, bioContent=null, aboutCorrespAuthor=null), CN=AuthorExt(id=1299828221986500883, tenantId=1146029695717560320, journalId=1296125453100220459, articleId=1297211746784141990, authorId=1299828221814534415, language=CN, stringName=陈以钒, firstName=null, middleName=null, lastName=null, prefix=null, suffix=null, authorComment=null, nameInitials=null, affiliation=null, department=null, xref=null, address=1华中农业大学工学院,武汉 430070
2农业农村部水产养殖设施工程重点实验室,武汉 430070, bio=null, bioImg=null, bioContent=null, aboutCorrespAuthor=null)}, companyList=[AuthorCompany(id=1299828221177000188, tenantId=1146029695717560320, journalId=1296125453100220459, articleId=1297211746784141990, xref=1, ext=[AuthorCompanyExt(id=1299828221185388797, tenantId=1146029695717560320, journalId=1296125453100220459, articleId=1297211746784141990, companyId=1299828221177000188, language=EN, country=null, province=null, city=null, postcode=null, companyName=null, departmentName=null, remark=1College of Engineering, Huazhong Agricultural University, Wuhan 430070, China), AuthorCompanyExt(id=1299828221193777406, tenantId=1146029695717560320, journalId=1296125453100220459, articleId=1297211746784141990, companyId=1299828221177000188, language=CN, country=null, province=null, city=null, postcode=null, companyName=null, departmentName=null, remark=1华中农业大学工学院,武汉 430070)]), AuthorCompany(id=1299828221269274879, tenantId=1146029695717560320, journalId=1296125453100220459, articleId=1297211746784141990, xref=2, ext=[AuthorCompanyExt(id=1299828221277663488, tenantId=1146029695717560320, journalId=1296125453100220459, articleId=1297211746784141990, companyId=1299828221269274879, language=EN, country=null, province=null, city=null, postcode=null, companyName=null, departmentName=null, remark=2The Key Laboratory of Aquaculture Facilities Engineering, Ministry of Agriculture and Rural Affairs, Wuhan 430070, China), AuthorCompanyExt(id=1299828221281857793, tenantId=1146029695717560320, journalId=1296125453100220459, articleId=1297211746784141990, companyId=1299828221269274879, language=CN, country=null, province=null, city=null, postcode=null, companyName=null, departmentName=null, remark=2农业农村部水产养殖设施工程重点实验室,武汉 430070)])]), Author(id=1299828222036832533, tenantId=1146029695717560320, journalId=1296125453100220459, articleId=1297211746784141990, orderNo=3, firstName=null, middleName=null, lastName=null, nameCn=null, orcid=null, stid=null, country=null, authorPic=null, dead=0, email=null, emailSecond=null, emailThird=null, correspondingAuthor=0, authorType=1, ext={EN=AuthorExt(id=1299828222124912920, tenantId=1146029695717560320, journalId=1296125453100220459, articleId=1297211746784141990, authorId=1299828222036832533, language=EN, stringName=Yiren ZHANG, firstName=Yiren, middleName=null, lastName=ZHANG, prefix=null, suffix=null, authorComment=null, nameInitials=null, affiliation=null, department=null, xref=null, address=1College of Engineering, Huazhong Agricultural University, Wuhan 430070, China
2The Key Laboratory of Aquaculture Facilities Engineering, Ministry of Agriculture and Rural Affairs, Wuhan 430070, China, bio=null, bioImg=null, bioContent=null, aboutCorrespAuthor=null), CN=AuthorExt(id=1299828222196216089, tenantId=1146029695717560320, journalId=1296125453100220459, articleId=1297211746784141990, authorId=1299828222036832533, language=CN, stringName=张义仁, firstName=null, middleName=null, lastName=null, prefix=null, suffix=null, authorComment=null, nameInitials=null, affiliation=null, department=null, xref=null, address=1华中农业大学工学院,武汉 430070
2农业农村部水产养殖设施工程重点实验室,武汉 430070, bio=null, bioImg=null, bioContent=null, aboutCorrespAuthor=null)}, companyList=[AuthorCompany(id=1299828221177000188, tenantId=1146029695717560320, journalId=1296125453100220459, articleId=1297211746784141990, xref=1, ext=[AuthorCompanyExt(id=1299828221185388797, tenantId=1146029695717560320, journalId=1296125453100220459, articleId=1297211746784141990, companyId=1299828221177000188, language=EN, country=null, province=null, city=null, postcode=null, companyName=null, departmentName=null, remark=1College of Engineering, Huazhong Agricultural University, Wuhan 430070, China), AuthorCompanyExt(id=1299828221193777406, tenantId=1146029695717560320, journalId=1296125453100220459, articleId=1297211746784141990, companyId=1299828221177000188, language=CN, country=null, province=null, city=null, postcode=null, companyName=null, departmentName=null, remark=1华中农业大学工学院,武汉 430070)]), AuthorCompany(id=1299828221269274879, tenantId=1146029695717560320, journalId=1296125453100220459, articleId=1297211746784141990, xref=2, ext=[AuthorCompanyExt(id=1299828221277663488, tenantId=1146029695717560320, journalId=1296125453100220459, articleId=1297211746784141990, companyId=1299828221269274879, language=EN, country=null, province=null, city=null, postcode=null, companyName=null, departmentName=null, remark=2The Key Laboratory of Aquaculture Facilities Engineering, Ministry of Agriculture and Rural Affairs, Wuhan 430070, China), AuthorCompanyExt(id=1299828221281857793, tenantId=1146029695717560320, journalId=1296125453100220459, articleId=1297211746784141990, companyId=1299828221269274879, language=CN, country=null, province=null, city=null, postcode=null, companyName=null, departmentName=null, remark=2农业农村部水产养殖设施工程重点实验室,武汉 430070)])])]
李亚苹 , 谭鹤群* , 陈以钒, 张义仁
作者信息
  • 1华中农业大学工学院,武汉 430070
  • 2农业农村部水产养殖设施工程重点实验室,武汉 430070
通讯作者:
谭鹤群,博士,教授,博士生导师,研究方向为畜禽水产养殖工程技术与装备。Email:
作者简介:

李亚苹,博士生,研究方向为现代养殖技术与装备。Email:

Dual-model progressive feeding strategy for Micropterus salmoides
Yaping LI , Hequn TAN* , Yifan CHEN, Yiren ZHANG
Affiliations
  • 1College of Engineering, Huazhong Agricultural University, Wuhan 430070, China
  • 2The Key Laboratory of Aquaculture Facilities Engineering, Ministry of Agriculture and Rural Affairs, Wuhan 430070, China
出版时间: 2026-06-30 doi: 10.11975/j.issn.1002-6819.202601127
文章导航
收藏切换

针对大口黑鲈养殖过程中难以精确且自动调整投喂量的现状,该研究提出了一种基于双模型渐进式的大口黑鲈投喂策略。采用单场多轮投喂方式,采集每轮投喂中包括音频在内的多模态数据。使用动态自适应阈值分割方法来检测单轮音频中大口黑鲈的摄食音频,基于提取的摄食音频特征进行主成分分析获取总摄食特征;使用特征选择算法筛选出大口黑鲈投喂状态的敏感特征作为模型的输入,用于学习和分类。结果表明,特征选择算法筛选出的大口黑鲈投喂状态的敏感特征为总摄食特征、光照强度、水温、体质量和单轮投喂率;对比了原始特征集与敏感特征集在梯度提升决策树和随机森林上的分类性能差异,结果证明了敏感特征的有效性;基于Transformer-RC的投喂状态预测模型在1~3时间窗口上的表现均优于其他三种模型,平均准确率为0.93~0.94;5种集成学习模型在投喂状态分类方面优于3种基础模型,其中基于LightGBM的投喂状态分类模型性能表现最佳,平均准确率为0.98;双模型渐进式投喂策略在饲料需求优化模型辅助下指导投喂,饵料系数较传统投喂降低约40%。该研究可为开发智能投喂系统提供依据。

水产养殖  /  大口黑鲈  /  精准投喂  /  双模型  /  投喂策略

Accurately adjusting the feeding intake is often required in pond aquaculture of Micropterus salmoides. In this study, a dual-model progressive feeding strategy was proposed to accurately forecast the postprandial feeding status of the subsequent round using a time series model. The decrement mechanism was first triggered. Subsequently, a classification model was utilized to determine the postprandial feeding status of the current round, enabling the termination of the decrement mechanism. The experimental system consisted of three pond culture tanks stocked with Micropterus salmoides, with initial body masses of (161.47±11.02) , (204.36±17.09) , and (220.25±22.78) g, and the stocking densities of 224, 301, and 294 individuals, respectively. Feeding experiments were conducted using a multi-round feeding protocol within a single session. Audio data was collected during feeding using the digital hydrophone, while video data was obtained using a camera. Meanwhile, multimodal data including light intensity, water temperature, body mass, per-round feeding rate, and population abundance were synchronously recorded for per-round feeding. The stabilization of the cumulative feeding energy curve was adopted to determine feeding termination. A dynamic adaptive threshold segmentation was employed to accurately identify the feeding audio within the collected audio. Principal component analysis was conducted on the feeding features to extract from the feeding audio. Feature selection was employed to identify sensitive features to feeding status, which served as the inputs for model training and classification. Four models 1D convolutional neural network, Long short-term memory, gated recurrent unit, and transformer were improved to construct the feeding prediction models with time series. The optimal model was selected to predict the postprandial feeding status of the subsequent round, serving as the first-stage model of feeding strategy. Eight models k-Nearest Neighbors, decision tree, support vector machine, random forests, adaBoost, gradient boosting decision trees, extreme gradient boosting, and light gradient boosting machine (LightGBM) were employed to construct feeding status classification models. The best classification performance was selected as the feeding determination model in the second stage. Results showed that the sensitive features included overall features, light intensity, water temperature, body mass, and per-round feeding rate. A classification between the original and sensitive features demonstrated the effectiveness of the sensitive features, according to gradient boosting decision trees and random forests models. Transformer-regression-classification (transformer-RC) model outperformed the rest of the models across 1 to 3-time windows, with the accuracy from 0.93 to 0.94. The Transformer-RC model effectively predicted the postprandial feeding status of the subsequent round. All five ensemble models achieved strong classification, which outperformed the three base models. Among them, the LightGBM model achieved an accuracy of 0.98 for inputs to the classification model of feeding status in the second stage of the feeding strategy. Both the Transformer-RC and LightGBM models shared high prediction and classification after validation, with average values of four evaluation metrics exceeding 0.89 under both feeding states, indicating strong generalization. This finding can provide a strong reference to develop intelligent feeding.

aquaculture  /  Micropterus salmoides  /  precision feeding  /  dual-model  /  feeding status
李亚苹, 谭鹤群, 陈以钒, 张义仁. 基于双模型渐进式的大口黑鲈投喂策略. 农业工程学报, 2026 , 42 (12) : 384 -396 . DOI: 10.11975/j.issn.1002-6819.202601127
Yaping LI, Hequn TAN, Yifan CHEN, Yiren ZHANG. Dual-model progressive feeding strategy for Micropterus salmoides[J]. Transactions of the Chinese Society of Agricultural Engineering, 2026 , 42 (12) : 384 -396 . DOI: 10.11975/j.issn.1002-6819.202601127
饲料是决定水产养殖产业成本的主要因素[1],不适当的投喂策略难以满足 “提质增效,绿色发展”的现代渔业发展需求。饲料投喂过少会影响鱼类生长速度,容易造成鱼体大小不一,甚至出现鱼之间互相残食的现象[2-3];投喂过量则会造成饲料浪费,减少了养殖利润,残饵会加剧水体污染[4-5]。目前,鱼类饲喂主要依靠人工经验和预设的投喂时间和投喂量,容易受到养殖人员的主观经验和意识的影响,不能根据养殖环境、鱼类的生长信息和鱼群摄食过程中的参数信息精确地自动调整投喂策略[6-8]。量化摄食需求和鱼类行为是解决上述问题的有效途径。
量化摄食需求技术主要基于生物能量的营养模型和水质参数。基于生物能量的营养模型允许根据鱼类预期的生长性能、营养需求和膳食营养成分确定投喂量[9],LIU等[10]基于生物能量学原理,构建了饲料需求模型,用于预测饲料需求量。水质参数是影响鱼类摄食的主要因素,这些参数可以间接反映鱼类的饥饿状态[11-12],CHEN等[13]以水温、溶解氧浓度、体质量及种群数量作为输入,基于反向传播神经网络和思维进化协同优化的计算框架构建了摄食量预测模型。两种量化摄食需求的方法均未考虑疾病等因素对鱼类摄食的影响,导致模型在病理条件下稳健性较差。事实上,大多数鱼类疾病或水质不适均会导致摄食量下降及摄食活动减弱,因此摄食行为可作为反映疾病及环境应激等因素的重要指标。量化摄食行为的技术主要包括机器视觉和声学。基于机器视觉的鱼类摄食行为量化方法可分为直接量化和间接量化两种[14-15]。直接量化主要通过分析鱼群的形状、纹理、运动速度等特征量化摄食强度[3, 6],CAI等[16]提出了一种用于鱼类摄食行为识别的双阶段框架,通过提取鱼群的空间分布信息,实现了鱼类摄食行为的快速且精准识别。间接量化主要通过检测残饵数量、水面反射及水花变化间接评估摄食行为[17],LI等[18]开发了一种自适应阈值检测方法,该方法在水浊度不同且照明不均的水下环境中,能够有效检测剩余鱼食。声学方法包括主动声学和被动声学。主动声学技术通过信号回波定位和监测鱼群,以此判断鱼群的瞬时游动速度与方向,进而评估当前摄食强度[19],ANRAS和LAGARDÈRE[20]利用声学定位遥测系统标记鱼类在不同觅食强度下游泳行为的差异,以分类觅食行为。被动声学技术通过采集鱼类咀嚼、游动及拍水产生的声音信号,识别和监测鱼类觅食行为[21],ZENG等[22]将摄食音频信号转换为声谱图用于摄食状态分类。基于机器视觉和声学的方法均属于事后分析,其摄食强度分类结果滞后于投喂过程,导致投喂终止指令的发出必然滞后于实际摄食需求,从而产生饲料浪费。
为了应对上述挑战,本研究采用单场多轮的投喂方式开展投喂试验,提出一种双模型渐进式投喂策略,主要包括两个子任务:预测下一轮摄食后对应的投喂状态和分类本轮投喂后对应的投喂状态。预测下一轮摄食后对应的投喂状态的目的是触发减量机制,若预测投喂状态为停止投喂(stop feeding,SF),表明按原定投喂量将导致饲料浪费。分类本轮投喂后对应的投喂状态的目的是终止减量机制,若投喂状态为SF,结束本场投喂。针对首个子任务,本研究将时序建模模块嵌入其网络架构,以增强模型对时序数据的自适应分析能力。目前水产领域基于时序研究集中于溶解氧预测,而摄食行为的时序预测仍是空白。CAO等[23]采用基于注意力机制的门控循环单元(gated recurrent unit,GRU)准确预测中心监测点的溶氧含量,结合位置信息和时间信息,构建基于RS-GBRT的溶氧含量三维预测模型,准确预测整个池塘溶氧的三维分布。GUO等[24]通过主成分分析(principal component analysis,PCA)提取影响大口黑鲈溶氧的关键因素,利用探路者算法(pathfinder algorithm,PFA)优化算法对GRU神经网络的关键参数进行自动优化,构建基于PCA-PFA-GRU的组合预测模型,预测大口黑鲈养殖水质中的溶氧量。机器学习凭借其高效的计算性能和卓越的分类精确率,在信息挖掘技术驱动的分类任务中展现出显著优势,并已广泛应用于农业生产领域。XU等[25]使用机器学习算法来构建适合处方大数据的农作物病害诊断模型,总体准确率为 80.36%。JIANG等[26]提出一种基于证据推理规则和堆叠集合学习的小麦品质精确分类模型,准确率达到 88.1%。这些成功案例为投喂状态的分类提供了新的思路。
本研究利用特征选择算法筛选出对大口黑鲈投喂状态敏感的特征,并使用原始数据集和经过特征筛选后的数据构建分类模型,以验证筛选敏感特征的有效性;对比4种基于时间序列的投喂状态预测模型,筛选出性能最优的模型作为投喂策略第一阶段的模型,用于触发减量机制;对比8种分类模型筛选出性能最优的模型作为第二阶段的投喂状态判别模型,用于终止减量机制。
试验在华中农业大学金口畜禽水产养殖试验基地(中国,武汉)进行。试验系统包括3个池塘圈养桶(直径为4 m,总深度为2.4 m),圈养桶中大口黑鲈的初始体质量分别为(161.47±11.02)、(204.36±17.09)、(220.25±22.78)g,养殖尾数分别为224、301、294尾。大口黑鲈已在池塘圈养桶中养殖超过1个月,认为已适应该环境下的饲料投喂。大口黑鲈每日投喂两场,时间分别为8:00和17:00。为保证大口黑鲈体质量数据的多样性,每5 d开展一次投喂试验,试验的前1 d称量体质量(body mass, BM),试验持续140 d。试验结束时三个圈养桶中大口黑鲈的体质量分别为(428.61±15.47)、(450.71±17.40)、(482.25±20.24 )g。因为鱼会死于疾病或其他原因,因此,应记录每次投喂试验对象的种群丰度(population abundance,PA)。
本研究采用单场多轮的投喂方式开展投喂试验,每轮投喂量共设置4个梯度,分别为100 、200 、300 、400 g,投喂前架设LST-DH01系列数字水听器(96KHZ,24 bit)和DS-2CD3T47DWD-L摄像机(分辨率为2560×1440像素,帧率为25帧/s)同步录制摄食音频和视频。数字水听器的水听器端置于池塘圈养桶的中心位置,位于水面以下0.3 m,另一端通过网线与计算机连接,大口黑鲈的强烈进食行为可能会损坏水听器,因此在其外部安装了一个保护笼。数字水听器开启之后录制3 min背景噪声用于降噪。随后按照设置的投喂量进行投喂,计算机实时接收数据并记录;确定水面无残余饵料并平静超过10 s继续投喂,当水面有明显残饵且大口黑鲈摄食行为终止时立即停止投喂。投喂期间,使用数字水听器收集大口黑鲈的摄食音频,摄像机收集摄食视频,同时记录由MDS-H11ISHP10-L05水温传感器(温度范围为0~60 ℃,分辨率为0.1 ℃,精确率为0.1 ℃)和JD-GZ光照强度传感器(量程为0~157 286 lx,精确度低于3%)监测的水温(water temperature, WT)与光照强度(light intensity, LI)。本文将单轮投喂量与大口黑鲈总质量的比值命名为单轮投喂率(per-round feeding rate, PFR)。鱼群摄食音视频信息采集系统如图1所示。
根据ØVERLI等[27]提出的鱼类摄食强度分类标准,本文将摄食强度分为“强”、“中”、“弱”、“无”4 类,如表1 所示。
本研究的投喂方式为单场多轮,因此为准确提取单轮投喂音频信号,需对采集的音频数据进行手工切分和标记,以构成试验所需音频信号。本研究通过声像同步的采集方法将大口黑鲈摄食过程中的音频信号与视频行为精准对应,因此在音频信号初步处理中采用Adobe Audition 2020软件中的多轨模式进行声像同步处理,结合音频波形特征与视频中的摄食行为识别并标记各轮次摄食信号,之后通过人工切分获得试验所需的单轮摄食音频样本。同时,采用鱼类摄食强度分类标准对单轮音频中大口黑鲈摄食后期的行为进行分类,继而根据分类结果对单轮音频进行编号。本研究共获取86场投喂的序列数据,切分出1302段音频。其中,继续投喂(continue feeding,CF)状态913段音频,SF状态389段音频。
音频采集过程中,岸边增氧设备的运行声音和圈养桶内碳纳米管曝气的声音均会对目标信号特征造成干扰。为分离出纯净的大口黑鲈摄食音频信号,本研究采用谱减法进行降噪,以提高后续摄食行为特征提取的准确性。该方法的核心思想是从含噪信号的幅度谱中减去噪声估计谱,从而增强目标信号。谱减法的步骤如下:
1) 带噪信号$ y(t) $由纯净信号$ x(t) $和噪声$ n(t) $组成
$ y(t)=x(t)+n(t) $
2) 短时傅里叶变换
$ Y(\omega ,k)=X(\omega ,k)+N(\omega ,k) $
式中Y(ω,k)、X(ω,k)和N(ω,k)分别表示带噪信号、纯净信号和噪声的短时傅里叶变换复数谱;ω为频率索引;k为第k帧。
3) 噪声估计
$ \left|\hat{N}\left(\omega\right)\right|^2=\frac{1}{M}\sum\limits_{k=1}^M\left|N(\omega,k)\right|^2 $
$ M=\frac{T-L+H}{H} $
式中$ \left|\hat{N}\left(\omega\right)\right|^2 $为估计的噪声功率谱;M为噪声帧数;T为噪声音频时长,ms;L为帧长,ms;H为帧移,ms。
4) 谱减法
$ \begin{split} & \left|\hat{X}\left(\omega,k\right)\right|^2= \\ &\left\{\begin{array}{llll}\left|Y(\omega,k)\right|^2-\alpha\cdot\left|\hat{N}\left(\omega\right)\right|^2 & \left|Y\left(\omega,k\right)\right|^2\ge\alpha\cdot\left|\hat{N}\left(\omega\right)\right|^2 \\ \beta\cdot\left|\hat{N}\left(\omega\right)\right|^2 & \left|Y\left(\omega,k\right)\right|^2 \lt \alpha\cdot\left|\hat{N}\left(\omega\right)\right|^2\end{array}\right.\end{split} $
式中$ \left|\hat{X}\left(\omega,k\right)\right|^2 $为估计的纯净信号功率谱;$ \alpha $为过减因子,β设置为0.002[28-29]
为了减小较大$ \alpha $值引起的语音失真,本研究采用基于信噪比(signal-to-noise ratio,SNR)的自适应方法计算$ \alpha $。其基本思想是:减法的幅度应依据每帧噪声信号的SNR动态调整,即在高SNR条件下减量较小,而在低SNR条件下减量较大。分段噪声信号SNR计算如下:
$ \mathrm{SNR}=10\mathit{\mathrm{\times1\mathrm{g}_{ }}}\frac{\left|\left|\left(Y(\omega,k\right)\right|\right|_2^2}{\left|\left|\hat{N}\left(\omega\right)\right|\right|_2^2} $
过减因子α计算如下:
$ \alpha =\left\{\begin{array}{*{20}{c}} 5, & {\mathrm{SNR}}< -5\;{\mathrm{db}}\\4-\dfrac{3}{20}\mathrm{}{\mathrm{SNR}}, & -5\;{\mathrm{db}}\leq {\mathrm{SNR}}\leq 20\;{\mathrm{db}}\\1, & {\mathrm{SNR}}> 20\;{\mathrm{db}}\end{array}\right. $
针对单轮音频信号持续时间较长导致计算复杂度高的问题,本研究采用分帧处理策略将连续音频信号转换为时序特征序列,设置帧长为100 ms,帧移为50 ms[30]。为了降低帧截断引起的频谱泄漏效应,选取窗口长度为100 ms的汉明窗对信号帧进行加权处理。窗口化后第k帧的信号如式(8)所示:
$ {y}_{k}(j) =w(j)\times x(H\times k+j) $
式中$ w\left(j\right)=0.54-0.46\mathrm{cos} \left(\dfrac{2\pi j}{N-1}\right) $N为汉明窗的长度;j为第k帧的第j个采样点。
本研究使用前期课题组开发的摄食端点检测算法[31],提取单轮音频信号中从开始摄食到结束摄食的信号,为后续的分析提供可靠的数据基础。信号处理系统搭建在PC端的Python软件中,通过Python软件编程对预处理后的信号进行端点检测。在单轮投喂前期,大口黑鲈的集群摄食行为导致累积摄食活动快速增加;随着水面饲料的消耗,摄食活动强度逐渐下降,直至累积能量曲线的斜率趋近于零,此时判定为摄食终止点[32]。因此,该算法以累计摄食能量曲线趋于平稳作为摄食结束的依据,进行摄食端点检测。摄食端点检测算法流程图如图2所示,端点检测的步骤如下:
1)获取随时间变化的短时能量;
2)根据短时能量计算累计能量,构建动态能量曲线;
3)对累计能量曲线进行平滑处理并求取其斜率。
4)对后8%能量对应的斜率进行阈值分割进而获取摄食端点。
短时能量$ E\left(t_k\right) $计算如下:
$ E\left({t}_{k}\right)=\sum\limits_{j=0}^{L-1}{({{y}_{k}}\left( j\right) )}^{2} $
式中$ t_k=H\times k+\dfrac{L}{2} $
累计能量$ E_{cum}\left(t_k\right) $计算如下:
$ {E}_{cum}\left({t}_{k}\right)=\sum\limits_{k=0}^{m}E({t}_{k}) $
式中m表示累计帧数。
累计能量的斜率$ S\left(t_k\right) $计算如下:
$ S\left({t}_{k}\right)=\frac{240}{W({W}^{2}-1)}\sum\limits_{f=-(W-1)/2}^{(W-1)/2}f\times {E}_{cum}({t}_{k+f}) $
式中$ t_{k+f}=H\times\left(k+f\right)+\dfrac{L}{2} $W为Savitzky-Golay 滤波器的窗口长度,取21;f为滤波器窗口内的偏移索引。
本研究使用动态自适应阈值分割算法对单轮音频信号的后8%能量段进行斜率分析,以实现摄食端点的精确检测。通过分析大口黑鲈摄食行为发现,摄食终止后因饥饿引发的非摄食活动仍会产生残余声学能量。针对这一现象,本研究通过统计分析并参考LI等[31]的研究选取300 V2作为初始分割阈值。
课题组前期研究发现,摄食音频的功率(power,PW)、过零率(zero-crossing rate,ZCR)、平均谱熵(mean spectral entropy,MSE)3个特征与摄食量均呈极显著相关(P < 0.001)。这些特征是基于整段摄食音频信号的全局统计量,能够有效描述信号的整体特性。PW是表征音频信号能量强度的时域特征,通过离散信号采样值的平方均值计算;ZCR是表征音频信号时域变化率的重要特征,反映信号波形在单位时间内穿过零电平的次数;MSE表征声学信号频谱能量分布均匀性的非线性动力学特征,表征信号的复杂度。PW、ZCR、MSE的计算如式(12)~(14)所示:
$ \mathrm{PW}=\frac{\displaystyle\sum\limits_{t=0}^{N-1}x^2_1(t_1)}{N_1} $
$ \mathrm{ZCR}=\frac{1}{2(N_1-1)}\sum\limits_{t=1}^{N_1-1}|\mathit{\mathrm{sgn}}\left[x\left(t\right)\right]-\mathit{\mathrm{sgn}}\left[x\left(t-1\right)\right]| $
$ \mathrm{MSE}=\frac{-\displaystyle\sum\limits_{h=0}^{N_1/2}P_t\left(h\right)\log_2[P_t\left(h\right)]}{N_1} $
式中 $ x_1(t_1) $ 是离散信号的第t个采样点;N1为总采样数;$ {\mathrm{sgn}}\left[x(t)\right]=\left\{\begin{array}{llllllllllll}-1, & (x(t) \lt 0) \\ 1, & (x(t)\ge0)\end{array}\right. $$ {P}_{t}\left(h\right) $是归一化的第h个频率分量的功率谱。
相关性分析表明,3个摄食特征之间存在极显著相关性(P<0.001)(表2)。高度相关的特征直接作为模型输入会导致多重共线性问题,引起模型对训练数据中噪声和异常值的过度拟合,从而损害模型的泛化能力。为解决这一问题,本研究采用PCA进行特征降维,得到少量互不相关的主成分作为模型的输入。这种处理不仅有效消除了特征间的冗余信息,同时最大程度地保留了原始数据的变异特征,从而提升模型的稳健性和预测性能。
主成分分析结果如表3所示。由表3可以看出,第一个主成分的特征值大于1,且主成分方差贡献率达到82.166%,其余主成分的特征值均远小于第一个主成分的特征值,因此,根据Kaiser准则选取第一个主成分代替原变量。按照主成分的方差贡献率构造综合评分公式计算总摄食特征(overall feeding feature,OFF),如式(15)所示。
$ \mathrm{OFF}=0.82166\times(0.363X_1-0.359X_2+0.380X_3) $
式中$ {X}_{1}~{X}_{3} $依次为PW、ZCR、MSE的标准分。
试验获取86场投喂的序列数据,共1302组数据。每组数据包含了BM、PA、PFR、WT、LI、OFF和投喂状态。归一化有助于加快损失函数的收敛速度,防止网络在训练过程中出现梯度爆炸,提高计算精确率,本研究采用式(16)[33]将输入数据归一化处理至[0,1]范围内。
$ x_i^*=\frac{x_i-x\mathrm{_{min}}}{x\mathrm{_{max}}-x\mathrm{_{min}}} $
式中$ {x}_{i} $为原始数据,$ x\mathrm{_{max}} $$ {x}_{min} $分别为原始数据中的最大值和最小值,$ x_{i}^{*} $为归一化之后的数据。
为了筛选出对大口黑鲈投喂状态敏感的变量,本研究选择梯度提升决策树(gradient boosting decision tree,GBDT)和随机森林(random forest,RF)进行特征筛选。这两种方法均属于嵌入式特征选择算法,能够结合模型训练过程自动评估特征重要性。在训练过程中,GBDT能够自动计算特征的重要性,从而实现对关键特征的筛选[34]。为了避免仅使用一种方法实现特征选择的偶然性,同时对比不同算法筛选大口黑鲈摄食状态敏感特征的结果,使用RF算法对特征进行筛选。RF通过构建多棵决策树并集成其结果,利用特征在树节点分裂时的不纯度减少来评估特征重要性[35],其随机子采样和特征随机选择的机制有助于降低过拟合风险,并能稳定地筛选出关键特征。
本研究提出了基于时间序列的投喂状态预测模型,用于在大口黑鲈的投喂中预测下一轮摄食后对应的投喂状态。该模型作为双模型渐进式的大口黑鲈投喂策略第一阶段的模型,其作用为触发投喂的减量机制。本研究分别对一维卷积神经网络(1D convolutional neural network,1D-CNN)、长短期记忆网络(long short-term memory,LSTM)、GRU、Transformer模型进行改进以构成基于时间序列的投喂状态预测模型,通过对4种基于时间序列的投喂状态预测模型性能的评估,选取大口黑鲈投喂状态的最佳预测模型。这些模型包含各种架构和注意力机制,使本研究具有广泛的适用性和代表性。
基于时间序列的投喂状态预测模型包括1D-CNN-RC、LSTM-RC、GRU-RC和Transformer-RC,这些模型可用于时序特征提取和多任务联合优化的复杂场景。前三种分类模型分别以1D-CNN、LSTM和GRU为特征提取网络,从数据集中提取相关特征。Transformer-RC以带有相对位置编码的Transformer编码器作为特征提取网络,相对位置编码提高了模型捕获长期依赖关系的能力,从而实现了更有效的时间特征表示。这些模型增加了回归层和分类层,使用全连接作为回归层,将提取的特征映射到目标维度;多层感知机(multilayer perceptron,MLP)作为模型的分类层(图3)。本研究提出的模型架构创新性地将回归层的输出特征与原始分类特征进行拼接,作为分类层的输入。这种端到端的设计模式显著提升了模型的整体性能和训练效率。
1D-CNN是专门用于处理一维序列数据的深度学习模型,基本架构包括卷积层、激活函数层、池化层和全连接层[36]图4)。本研究所使用的1D-CNN网络由两层一维卷积构成(输入通道为5,输出通道分别为128和256,卷积核均为3,填充为1),每层卷积后依次连接批归一化、ReLU和Dropout(设置为0.2),以提取序列特征并抑制过拟合,卷积输出通过自适应平均池化后作为回归层输入。
LSTM 是一种特殊的循环神经网络(recurrent neural network,RNN),能够有效学习序列数据中的长期短期依赖问题,是处理时序依赖任务的经典选择[37]。其核心思想是通过门控机制控制信息的流动,选择性保留或遗忘历史信息。该网络的结构由遗忘门、输入门和输出门组成(图5),输入门用于判断当前时步的输入是否应进入内存单元;遗忘门用于判断是否应遗忘前一个时间步的内存;输出门用于判断当前时步的内存是否传递给后一个时间步。LSTM 的循环结构允许它按顺序处理数据,并在每个时间步更新其内部状态。这种机制使 LSTM 能够自然地对时间序列数据中的时间依赖关系进行建模。LSTM 通过细胞状态和门控机制,允许梯度在长时间步中稳定传播,避免传统 RNN 的梯度消失和爆炸问题。LSTM 在小数据、实时系统等场景中具有重要价值。本研究所使用的LSTM网络以5维序列为输入,经过两层64单元的LSTM提取时序特征,最后时间步的隐藏状态作为回归层输入。
GRU是一种RNN,旨在处理序列数据。GRU在LSTM的基础上进行了简化,通过减少门控数量和合并细胞状态与隐藏状态,实现了更高的计算效率,同时保留其处理长期依赖的能力。该模型使用重置门和更新门控制信息的流动(图6)。重置门决定了如何将新的输入信息与前面的记忆相结合;更新门用于控制前一时刻的状态信息被带入到当前状态中的程度。模型的门控机制有效解决了传统RNN存在的梯度消失和梯度爆炸问题[38]。GRU的设计目的是在保持计算效率的同时,拥有较高的性能,适用于广泛的序列处理任务。本研究所使用的GRU网络以5维序列为输入,经过两层64单元的GRU提取时序特征,最后时间步的隐藏状态作为回归层输入。
Transformer 是一种基于自注意力机制的深度学习模型架构,其核心采用编码器-解码器框架实现序列到序列的建模。编码器负责将输入序列转换为一组特征表示,而解码器则利用这些特征生成输出序列。标准Transformer编码器包括前馈神经网络和多头注意力机制等两个主要组件(图7)。Transformer 采用多头注意力结构,增强了捕获长期依赖关系的能力。这种注意力机制通过动态调节权重分布来促进显著信息的优先级,从而确保在不同场景中选择相关特征[39-40]。本研究所使用的Transformer网络以5维序列为输入,经过线性嵌入和相对位置编码后,输入3层共享Transformer编码器(每层2个注意力头,前馈维度64,Dropout 0.1)提取时序特征,并将编码器最后时间步的输出作为回归层输入。
本研究提出了一种双模型渐进式的大口黑鲈投喂策略,特别关注大口黑鲈投喂状态的精确识别。该策略的第二阶段需要一个能够对本轮摄食后对应的投喂状态进行准确分类的模型。本研究选取k-近邻算法(k-nearest neighbors, KNN)、决策树(decision tree,DT)、支持向量机(support vector machine,SVM)、RF、自适应提升(adaptive boosting,adaboost)、GBDT、极限梯度提升算法(extreme gradient boosting,XGBoost) 和轻量级梯度提升机算法(light gradient boosting machine,LightGBM)等8种机器学习(machine learning,ML)算法建立大口黑鲈投喂状态分类模型,通过对模型分类性能的评估,选取最佳的模型作为第二阶段的投喂状态判别模型,用于终止减量机制。
KNN、DT和SVM是三种基础的机器学习分类模型。KNN是一种基于实例的惰性学习算法,通过计算样本与邻近训练数据的距离进行分类,简单直观但对数据规模和特征缩放敏感[41];DT通过树形结构对特征空间进行规则划分以实现分类,具有较强的可解释性,能够处理非线性特征,但容易发生过拟合;SVM则通过寻找最优超平面最大化分类间隔,尤其适合高维和小样本数据,对噪声和过拟合有较强鲁棒性,但计算复杂度较高[42]
RF、adaBoost、GBDT、XGBoost和LightGBM都属于集成学习方法,但它们的核心思想和优化策略不同,主要分为两大类:基于Bagging的集成方法和基于Boosting的集成方法。RF是基于Bagging的集成学习方法,通过Bootstrap抽样和特征子集选择构建多样化的决策树集合,并采用投票或平均机制进行最终预测[43]。AdaBoost、GBDT、XGBoost和LightGBM是基于Boosting的集成方法。AdaBoost通过自适应调整样本权重来聚焦误分类样本[44];GBDT基于梯度下降框架迭代拟合伪残差[45];XGBoost在GBDT基础上引入正则化项和二阶泰勒展开等创新;LightGBM则通过直方图算法、单边梯度采样和互斥特征捆绑等技术优化大规模数据处理[46]
本研究使用准确率(Accuracy)、精确率(Precision)、召回率(Recall)和F1分数(F1)等4个评价指标来评价模型的性能,分别由式(17)~(20)[47]所示。准确率是所有样本中被正确预测的比例。精确率是模型预测为正类的样本中实际为正类的比例。召回率是实际为正类的样本中被模型正确预测为正类的比例。F1分数是精确率和召回率的调和平均数。当准确率值接近1时,该算法具有更高的正确识别率。
$ \mathrm{Accuracy=\frac{TP+TN}{TP+TN+FP+FN}} $
$ \mathrm{Precision=\frac{TP}{TP+FP}} $
$ \mathrm{Recall=\frac{TP}{TP+FN}} $
$ F1=\mathrm{\frac{2TP}{2TP+FP+FN}} $
式中TP代表真阳性,TN代表真阴性,FP代表假阳性,FN代表假阴性。
基于时间序列的投喂状态预测模型在进行数据训练和预测时,回归任务的每组数据表示为$ (\boldsymbol{X_a},\boldsymbol{x_{a+n,b}}) $,其中$ \boldsymbol{X_a}=\left(\begin{matrix}\boldsymbol{x_{a,1}} & \cdots & \boldsymbol{x_{a,b}} \\ \vdots & \ddots & \vdots \\ \boldsymbol{x_{a+w-1,1}} & \cdots & \boldsymbol{x_{a+w-1,b}}\end{matrix}\right) $是输入特征矩阵,包含$ b $个特征参数,时间窗口长度为w$ {x}_{a+w,b} $为回归输出,表示第a+w轮的OFF;分类任务的每组数据表示为$ ({X}_{a},{y}_{a}) $,其中$ \boldsymbol{X_a}=\left(\boldsymbol{x_{a+w,1}} ,\cdots{{,}} \boldsymbol{x_{a+w,b}}\right) $是输入特征,$ {y}_{a} $是第a+w轮摄食后对应的投喂状态。投喂状态分类模型在进行数据训练和预测时,每组数据表示为$ (\boldsymbol{X_c},{y_c}) $,其中$ \boldsymbol{X_c}=\left(\boldsymbol{x_{c,1}} \mathrm{,}\cdots\mathrm{,} \boldsymbol{x_{c,b}}\right) $是输入特征向量,包含$ b $个特征参数,$ {y}_{c} $是第c轮投喂后对应的投喂状态。
本研究使用 Python 3.12.4 作为编程语言。数据集的80%用于训练模型,剩下的20%用于测试模型。为了优化模型性能,采用网格搜索和5折交叉验证对模型进行超参数调优[48]:在给定的参数范围内遍历所有可能的组合,利用5折交叉验证计算每组参数在验证集上的准确率、精确率、召回率和F1分数,并选择表现最佳的参数组合作为模型的最优超参数,从而保证模型在未见数据上的稳健性与泛化能力。本研究以准确率、精确率、召回率和F1分数作为评价指标,在测试集上评估基于时间序列的投喂状态预测模型和投喂状态分类模型的性能。建模过程如图8所示。
对BM、PA、PFR、WT、LI和OFF分别使用GBDT和RF算法进行大口黑鲈摄食状态敏感特征的筛选。GBDT算法设置“n_estimators”为65,“learning_rate”为0.09,“max_depth”为3,“min_samples_leaf”为5,“subsample”为0.7,“max_features”为“sqrt”;RF算法设置“n_estimators”为50,“max_depth”为8,“min_samples_leaf”为3,“max_features”为“sqrt”。2种算法得到各特征的敏感性排序如图9所示。图9a 为GBDT算法得到的各特征的重要性排序,依次为:OFF(0.6745)、LI(0.1895)、BM(0.0443)、WT(0.0427)、PFR(0.0402)和PA(0.0088)。图9b为RF算法得到的各特征重要性排序,依次为:OFF(0.5838)、LI(0.2158)、WT(0.0651)、PFR(0.0640)、BM(0.0591)和PA(0.0121)。在2种算法中重要性排序虽略有不同,但PA均为最不重要的特征。选择2种算法给出的重要程度排名前五的特征作为大口黑鲈摄食状态敏感特征以供后续分析,5个特征分别为:OFF、LI、WT、PFR、BM。
为了评估筛选特征的有效性,本研究对比了原始特征集与筛选后特征集在GBDT和RF上的分类性能差异。使用原始特征训练的模型在测试集中的混淆矩阵如图10a 所示,使用筛选后敏感特征训练的模型在测试集中的混淆矩阵如图10b 所示。混淆矩阵可视化了模型性能,其对角线表示正确识别投喂状态的置信度。深色表示高置信度,而浅色表示其他置信度较低的错误识别状态。可以看出,以筛选出的5个特征作为敏感特征构建的GBDT和RF模型同样可以在大口黑鲈摄食状态的分类中取得较好的效果。
为了具体分析所构建的分类模型在样本中的表现,将测试结果汇总得到表4。对比通过原始特征建立的摄食状态分类模型,通过敏感特征(剔除PA)建立的GBDT分类模型的验证集准确率、精确率、召回率和F1分数分别了上升了0.8%、0.5%、2.1%、1.6%;通过敏感特征建立的RF分类模型的验证集准确率、精确率、召回率和F1分数分别了上升了0.4%、0.5%、1.6%、0.5%。该结果证明了通过GBDT和RF算法筛选出的敏感特征的有效性,模型有效性和鲁棒性不受种群规模变化的影响。
本文对提出模型的不同时间窗口进行了测试,采用准确率、精确率、召回率和F1分数对模型在不同时间窗口上的表现进行评估(表5)。由表5可以看出,当时间窗口为1时,4种模型的综合性能由高到低排序为:Transformer- RC、LSTM- RC、1D-CNN- RC、GRU- RC,但该时间窗口下1D-CNN- RC和GRU- RC在SF类别上的精确率、召回率和F1分数存在低于0.90的情况;当时间窗口为2时,Transformer- RC模型的综合性能高于其余三种模型,但该时间窗口下4种模型在SF类别上的精确率和F1分数也存在低于0.90的情况;当时间窗口为3时,4种分类模型的综合性能由高到低排序为:Transformer- RC、GRU- RC、LSTM- RC/1D-CNN-RC。
综上所述,Transformer- RC在1~3时间窗口下的表现均最优,可能原因为Transformer内部的注意力机制能有效捕捉不同位置之间的距离依赖关系,更好地理解上下文和全局信息。1时间窗口对应的平均准确率、精确率、召回率和F1分数分别为0.94、0.935、0.93、0.935;2时间窗口对应的平均准确率、精确率、召回率和F1分数分别为0.93、0.92、0.94、0.93;3时间窗口对应的平均准确率、精确率、召回率和F1分数分别为0.94、0.94、0.94、0.94。这些性能指标表明,Transformer- RC模型可以有效预测下一轮摄食后对应的投喂状态。
此外,4种指标在Transformer- RC模型不同时间窗口下的综合表现按照从高到低排序为:3时间窗口 >1时间窗口 >2时间窗口。因此,在实际应用中,当投喂轮数<3时,采用1时间窗口模型作为第一阶段投喂状态预测模型;当投喂轮数≥3轮时,采用3时间窗口模型作为第一阶段投喂状态预测模型,以提升投喂状态预测的准确性与稳定性。
本文采用精确率、准确率、召回率和F1分数4种评价指标比较8种ML分类模型的性能(表6)。8种模型在类别SF上的分类性能均低于类别CF,这种性能差异可能与数据集的类别分布不平衡有关。基础模型在类别SF上的分类性能较差,KNN在类别SF上的精确率和F1分数均小于0.90,DT在类别SF上的精确率、召回率和F1分数均小于0.90。对于不平衡的数据集,集成学习模型表现出突出的优势。集成学习模型在类别SF上的分类性能与类别CF差异普遍较小,且5种集成学习模型在两种类别上的分类性能都较好。这是因为集成模型通过结合多个弱模型的预测结果,有效降低了单一模型的偏差和方差,从而提升了整体预测能力。总体来看,5种集成学习模型的性能均优于3种基础模型,平均准确率、精确率、召回率和F1分数均高于0.95,其中LightGBM的分类性能略优于其余4种模型,平均准确率、精确率、召回率和F1分数分别为0.98、0.965、0.98和0.97。
综合4种评价指标,本研究选取LightGBM模型作为大口黑鲈投喂策略第二阶段投喂状态的判别模型。
为进一步评估Transformer- RC和LightGBM模型的预测能力,在池塘圈养系统中(图11)开展了验证试验,共获得 321组数据。验证试验按照第1.1章节所述方法进行。在试验过程中,同步获取BM、PFR、WT、LI和音频数据。使用动态自适应阈值分割方法来检测单轮音频中大口黑鲈的摄食音频,基于提取的摄食音频特征进行主成分分析获取OFF。
表7 所示,本研究构建的Transformer-RC和LightGBM模型在321组数据上均表现出良好的预测与分类性能。在两种投喂状态下,四项评价指标的平均值均不低于0.89,表明所构建模型具有较强的泛化能力。与测试集结果相比,验证试验中模型的整体性能略有下降,究其原因可能为实际养殖环境具有较强的多变性,导致验证数据与训练和测试数据在分布上存在一定差异,从而对模型性能产生了一定影响。
本研究提出了一种双模型渐进式投喂策略,该策略分两阶段进行,第一阶段通过基于时间序列的投喂状态预测模型准确预测下一轮摄食后对应的投喂状态,用于触发减量机制;随后进入第二阶段,通过投喂状态分类模型来判别本轮摄食后对应的投喂状态,用于终止减量机制,实现更精准的投喂。在此基础上,双模型渐进式投喂策略与本课题组构建的饲料需求优化模型相结合,以生成最终的投喂决策输出,在保障鱼类福利的同时又能减少饲料浪费。投喂决策程序运行流程如图12所示,程序首先调用饲料需求优化模型生成理论投喂量M0,随后启动投喂流程。在初始投喂阶段,系统结合累计投喂量与摄食状态预测结果进行综合决策,当满足理论投喂量约束时终止投喂,当预测结果表明下一轮摄食后对应的投喂状态为“SF”时,进入减量投喂循环。在后续投喂阶段,系统结合累计投喂量与摄食状态分类结果进行综合决策,当满足理论投喂量约束或判别结果显示本轮摄食后对应的投喂状态为“SF”时结束投喂。
为验证投喂决策的实际效果,本研究在华中农业大学金口试验基地的循环水系统中开展了大口黑鲈投喂对比试验。试验选取两种规格的大口黑鲈开展投喂性能试验,每种规格均设置对照组和试验组。试验开始与结束时,每组均随机抽取30%的大口黑鲈进行质量测定,并计算平均值。试验期间,试验组均采用本文的投喂决策进行投喂,对照组采用传统投喂模式,投喂策略根据人工经验设定。除投喂模式不同外,各养殖桶日常养殖管理均保持一致,人工逐日统计养殖桶的总饲料量。
为评估不同投喂方式对大口黑鲈(Micropterus salmoides)生长状况的影响,选取饵料系数(feed conversion ratio,FCR)、增重率(weight gain ratio,WGR)和特定生长率(specific growth rate,SGR)作为衡量大口黑鲈生长状况的主要评价指标。其中,FCR表示单位BM增长所消耗的饲料量,用于反映饲料转化利用效率,其数值越低,说明饲料利用效率越高。WGR表示在一段时间内鱼体总质量的增长百分比,用于衡量该阶段鱼群质量的增加情况。SGR用于表征单位时间内鱼类的生长状况,其数值越大,表明单位时间内个体生长速率越快。FCR、WGR和SGR的计算方法分别如式(21)~(23)所示。
$ \mathrm{FCR}=\frac{F}{M_2-M_1} $
$ \mathrm{WGR\mathrm{ }}=\frac{M_2-M_1}{M_1}\times100\text{%} $
$ \mathrm{SGR}=\frac{\ln M_2-\ln M_1}{t_1}\times100\text{%} $
式中F为试验期间的总投喂量,kg;M1M2分别表示投喂试验开始和结束时大口黑鲈鱼群的总质量,kg;t1为养殖天数,d。
不同投喂模式下大口黑鲈的生长参数如表8所示,结果表明,对于小规格鱼,与对照组相比,试验组的WGR和SGR分别降低了17.84%和16.67%,同时FCR也降低了41.35%。这可能与起始BM差异有关,在增重相同的条件下,起始质量越大,WGR和SGR就越小。对于大规格鱼,与对照组相比,试验组的FCR降低了40.42%,同时WGR和SGR分别提高了28.13%和26.92%,综上所述,本研究使用的投喂决策在不同规格大口黑鲈养殖过程中均表现出良好的应用效果,能够在降低饲料系数的同时兼顾甚至提升生长性能,具有较高的实际应用价值与推广潜力。
本文提出的方法旨在实现集约化水产养殖系统内大口黑鲈的精准投喂,该方法结合饲料需求优化模型和双模型渐进式投喂策略,动态调整投喂量,逐步逼近摄食欲望。相比之下,传统的摄食行为识别模型属于事后分类,仍存在饲料浪费的情况。MA等[11]提出的TFFormer模型在鱼类摄食行为分类任务中实现了91.52%的准确率;UBINA等[14]通过构建两个深度卷积神经网络,分别用于光流估计和鱼类觅食强度评估,模型识别准确率达到95%;CAI等[16]提出了识别鱼类觅食行为的两阶段框架,在最小模型参数和计算约束下,准确率达到了83.33%;ZENG等[22]通过改进后的ASST网络实现了鱼类觅食行为的定量识别,准确率达到了96.16%。这些学者构建的模型在摄食行为分类方面虽然表现出较高的分类性能,但其研究未进一步评估投喂效能。张镇府等[49]构建的鲈鱼摄食状态分类模型对应的FCR为1.42,与本文结果相近;赵思琪等[50]提出的池塘养殖的精准投喂系统,其FCR为1.73,远远高于本文结果,表明本文方法在降低饵料消耗方面具有显著优势。本研究以大口黑鲈为例进行投喂策略研究,由于不同鱼类食性等方面的差异,该策略在评估其他鱼类摄食强度方面的表现仍有待验证。未来的研究将以自动化的精准投喂为目标,以饲料需求优化模型和双模型渐进式投喂策略为基础,结合自动投喂执行设备,开发一套水产养殖智能投喂管控系统。
1)通过特征选择算法筛选出的大口黑鲈投喂状态的敏感特征为总摄食特征、光照强度、水温、单轮投喂率和体质量,基于敏感特征的大口黑鲈分类模型的准确率、精确率、召回率和F1分数较基于原始特征的模型分别上升0.4%~0.8%、0.5%、1.6%~2.1%、0.5%~1.6%,证明了所筛选敏感特征的有效性;与LSTM- RC 、1D-CNN- RC 、GRU- RC等模型相比,Transformer-RC模型在1~3时间窗口上的表现均最优,准确率为0.93~0.94,可作为大口黑鲈投喂策略第一阶段的投喂状态预测模型;与3种基础模型相比,5种集成学习模型在大口黑鲈投喂状态分类上展现了明显的优势,其中基于LightGBM的投喂状态分类模型性能表现最佳,平均准确率为0.98,可作为大口黑鲈投喂策略第二阶段投喂状态的判别模型;验证试验表明,在两种投喂状态下四项评价指标的平均值均不低于0.89,证明了Transformer-RC和LightGBM模型较强的泛化能力;投喂对比试验表明,基于饲料需求优化模型与双模型渐进式投喂策略相结合的投喂方式饵料系数较传统投喂降低约40%。
2)本研究提出的双模型渐进式投喂策略首次将状态预测与动态调控相结合,在渐进式投喂过程中不断调整投喂量,从而实现了更科学、更高效的精准投喂控制。为提升模型的环境适应性,将水温和光照强度整合至特征工程框架中,实现了环境干扰的量化补偿,使得模型在干扰场景中表现出卓越的稳定性。

参考文献 引证文献
排序方式:
1
LI D L, WANG Z H, WU S Y, et al. Automatic recognition methods of fish feeding behavior in aquaculture: A review[J]. Aquaculture, 2020, 528: 735508.
2
WU T H, HUANG Y I, CHEN J M. Development of an adaptive neural-based fuzzy inference system for feeding decision-making assessment in silver perch (Bidyanus bidyanus) culture[J]. Aquacultural Engineering, 2015, 66: 41-51.
3
ZHOU C, ZHANG B H, LIN K, et al. Near-infrared imaging to quantify the feeding behavior of fish in aquaculture[J]. Computers and Electronics in Agriculture, 2017, 135(C): 233-241.
4
BARRAZA-GUARDADO R H, MARTÍNEZ-CÓRDOVA L R, ENRÍQUEZ-OCAñA L F, et al. Effect of shrimp farm effluent on water and sediment quality parameters off the coast of Sonora, México[J]. Ciencias Marinas, 2014, 40(4): 221-235.
5
ZHOU C, YANG X T, ZHANG B H, et al. An adaptive image enhancement method for a recirculating aquaculture system[J]. Scientific Reports, 2017, 7: 6243.
6
LIU Z, LI, X, FAN L Z, et al. Measuring feeding activity of fish in RAS using computer vision[J]. Aquacultural Engineering, 2014, 60: 20-27.
7
ZHOU C, LIN K, XU D M, et al. Near infrared computer vision and neuro-fuzzy model-based feeding decision system for fish in aquaculture[J]. Computers and Electronics in Agriculture, 2018, 146: 114-124.
8
ZHOU C, XU D M, LIN K, et al. Intelligent feeding control methods in aquaculture with an emphasis on fish: A review[J]. Reviews in Aquaculture, 2017, 10(4): 975-993.
9
DAVIS D A, HARDY R H. Feeding and Fish Husbandry[M]. London: Academic Press, 2022.
10
LIU X J, DU K H, ZHANG C Y, et al. Precision feeding system for largemouth bass (Micropterus salmoides) based on multi-factor comprehensive control[J]. Biosystems Engineering, 2023, 227: 195-216.
11
MA P C, YANG X T, HU W C, et al. Fish feeding behavior recognition using time-domain and frequency-domain signals fusion from six-axis inertial sensors[J]. Computers and Electronics in Agriculture, 2024, 227(2): 109652.
12
STONER A W, OTTMAR M L, HURST T P. Temperature affects activity and feeding motivation in Pacific halibut: Implications for bait-dependent fishing[J]. Fisheries Research, 2006, 81(2-3): 202-209.
13
CHEN L, YANG XT, SUN CH, et al. Feed intake prediction model for group fish using the MEA-BP neural network in intensive aquaculture[J]. Information Processing in Agriculture, 2020, 7(2): 261-271.
14
UBINA N M, CHENG S-C, CHANG C-C, et al. Evaluating fish feeding intensity in aquaculture with convolutional neural networks[J]. Aquacultural Engineering, 2021, 94: 102178.
15
WANG Y Q, YU X N, LIU J C, et al. Dynamic feeding method for aquaculture fish using multi-task neural network[J]. Aquaculture, 2022, 551: 737913.
16
CAI K W, YANG Z P, GAO T Y, et al. Efficient recognition of fish feeding behavior: A novel two-stage framework pioneering intelligent aquaculture strategies[J]. Computers and Electronics in Agriculture, 2024, 224: 109129.
17
LIU H Y, XU L H, LI D W. Detection and recognition of uneaten fish food pellets in aquaculture using image processing[C]. Sixth International Conference on Graphic and Image Processing (ICGIP), 2015, 9443: 94430G.
18
LI D W, XU L H, LIU H Y. Detection of uneaten fish food pellets in underwater images for aquaculture[J]. Aquacultural Engineering, 2017, 78(Part B): 85-94.
19
HUNG C C, TSAO S C, HUANG K H, et al. A highly sensitive underwater video system for use in turbid aquaculture ponds[J]. Scientific Reports, 2016, 6: 31810.
20
ANRAS M B, LAGARDÈRE J P. Measuring cultured fish swimming behaviour: first results on rainbow trout using acoustic telemetry in tanks[J]. Aquaculture, 2004, 240(1/2/3/4): 175-186.
21
DU Z Z, CUI M, WANG Q, et al. Feeding intensity assessment of aquaculture fish using Mel Spectrogram and deep learning algorithms[J]. Aquacultural Engineering, 2023, 102: 102345.
22
ZENG Y H, YANG X T, PAN L, et al. Fish school feeding behavior quantification using acoustic signal and improved Swin Transformer[J]. Computers and Electronics in Agriculture, 2023, 204: 107580.
23
CAO X K, REN N, TIAN G L, et al. A three-dimensional prediction method of dissolved oxygen in pond culture based on Attention-GRU-GBRT[J]. Computers and Electronics in Agriculture, 2021, 181: 105955.
24
GUO J J, DONG J Q, ZHOU B, et al. A hybrid model for the prediction of dissolved oxygen in seabass farming[J]. Computers and Electronics in Agriculture, 2022, 198: 106971.
25
XU C, DING J Q, QIAO Y, et al. Tomato disease and pest diagnosis method based on the Stacking of prescription data[J]. Computers and Electronics in Agriculture, 2022, 197: 106997.
26
JIANG H W, ZHANG S L, YANG Z, et al. Quality classification of stored wheat based on evidence reasoning rule and stacking ensemble learning[J]. Computers and Electronics in Agriculture, 2023, 214: 108339.
27
ØVERLI Ø, SøRENSEN C, NILSSON G E. Behavioral indicators of stress-coping style in rainbow trout: Do males and females react differently to novelty?[J]. Physiology & Behavior, 2006, 87(3): 506-512.
28
UPADHYAY N, KARMAKAR A. An improved multi-band spectral subtraction algorithm for enhancing speech in various noise environments[J]. Procedia Engineering, 2013, 64: 312-321.
29
BEROUTI M, SCHWARTZ R, MAKHOUL J. Enhancement of speech corrupted by acoustic noise[C]. IEEE International Conference on Acoustics, Speech, and Signal Processing (ICASSP). Washington, USA: IEEE, 1979: 208-211.
30
曹晓慧, 刘晃, 戚仁宇, 等. 循环水养殖大口黑鲈摄食颗粒饲料的声学特征[J]. 农业工程学报, 2021, 37(20): 219-225.
CAO Xiaohui, LIU Huang, QI Renyu, et al. Acoustic characteristics of the feeding pellets for Micropterus salmoides in circulating aquaculture[J]. Transactions of the Chinese Society of Agricultural Engineering (Transactions of the CSAE), 2021, 37(20): 219-225. (in Chinese with English abstract)
31
LI Y P, TAN H Q, ZHANG Y R, et al. A quantitative method of feeding demand for M. salmoides based on multi-source information fusion[J]. Biosystems Engineering, 2026, 262: 104372.
32
MALLEKH R, LAGARDÈRE J P, ENEAU J P, et al. An acoustic detector of turbot feeding activity[J]. Aquaculture, 2003, 221(1/2/3/4): 481-489.
33
KAVEH M, SHARABIANI V R, CHAYJAN R A, et al. ANFIS and ANNs model for prediction of moisture diffusivity and specific energy consumption potato, garlic and cantaloupe drying under convective hot air dryer[J]. Information Processing in Agriculture, 2018, 5(3): 372-387.
34
DELGADO-PANADERO A, HERNÁNDEZ-LORCA B, GARCÍA-ORDÁS M. T, et al. Implementing local-explainability in gradient boosting trees: Feature contribution[J]. Information Sciences, 2022, 589: 199-212.
35
DHILLON R, TAKOO G, SHARMA V, et al. Utilizing machine learning framework to evaluate the effect of climate change on maize and soybean yield[J]. Computers and Electronics in Agriculture, 2024, 221: 108982.
36
KIRANYAZ S, INCE T, ABDELJABER O, et al. 1-D Convolutional Neural Networks for Signal Processing Applications[C]// IEEE International Conference on Acoustics, Speech and Signal Processing (ICASSP), Brighton, UK: IEEE, 2019, 8360-8364.
37
MENG J Q, LI C B, TAO J, et al. RNN- LSTM-based model predictive control for a corn-to-sugar process[J]. Processes, 2023, 11(4): 1080.
38
李雯, 施珮. 融合特征信息和改进BiGRU的池塘养殖溶解氧预测模型[J]. 农业工程学报, 2025, 41(19): 218-226.
LI Wen, SHI Pei. Dissolved oxygen prediction model for pond culture by integrating feature information and improved BiGRU[J]. Transactions of the Chinese Society of Agricultural Engineering (Transactions of the CSAE), 2025, 41(19): 218-226. (in Chinese with English abstract)
39
LI Y X, JIA W X, WANG J M, et al. ALeRT-COVID: attentive lockdown-awaRe transfer learning for predicting COVID-19 pandemics in different countries[J]. Journal of Healthcare Informatics Research, 2021, 5: 98-113.
40
LIU Y Q, ZHANG Q, SONG L H, et al. Attention-based recurrent neural networks for accurate short-term and long- term dissolved oxygen prediction[J]. Computers and Electronics in Agriculture, 2019, 165: 104964.
41
COVER T, HART P. Nearest neighbor pattern classification[J]. IEEE Transactions on Information Theory, 1967, 13(1): 21-27.
42
BOSER B E, GUYON I, VAPNIK V N. A training algorithm for optimal margin classifiers[C]// Proceedings of the fifth annual workshop on Computational learning theory. New York, USA: Association for Computing Machinery, 1992, 144-152.
43
BREIMAN L. Random forests[J]. Machine Learning, 2001, 45: 5-32.
44
FREUND Y, SCHAPIRE R E. Experiment with a New Boosting Algorithm[C]// Proceedings of the Thirteenth International Conference on Machine Learning. San Francisco, CA, USA: Morgan Kaufmann Publishers Inc. , 1996, 148-156.
45
FRIEDMAN J H. Greedy Function Approximation: A gradient boosting machine[J]. Annals of Statistics, 2001, 29: 1189-1232.
46
KE G L, MENG Q, FINLEY T, et al. LightGBM: A highly efficient gradient boosting decision tree[C]// 31st Conference on Neural Information Processing Systems (NIPS). Red Hook, NY, USA: Curran Associates, Inc. , 2017.
47
YIN X Q, WU D H, SHANG Y Y, et al. Using an efficientnet-LSTM for the recognition of single cow’s motion behaviours in a complicated environment[J]. Computers and Electronics in Agriculture, 2020, 177: 105707.
48
JIANG P, CHEN J J. Displacement prediction of landslide based on generalized regression neural networks with K-fold cross-validation[J]. Neurocomputing, 2016, 198(19): 40-47.
49
朱明, 张镇府, 黄凰, 等. 基于轻量级神经网络MobileNetV3-Small的鲈鱼摄食状态分类[J]. 农业工程学报, 2021, 37(19): 165-172.
ZHU Ming, ZHANG Zhenfu, HUANG Huang, et al. Classification of perch ingesting condition using lightweight neural network MobileNetV3-Small[J]. Transactions of the Chinese Society of Agricultural Engineering (Transactions of the CSAE), 2021, 37(19): 165-172. (in Chinese with English abstract)
50
赵思琪, 赵三琴, 国振淇, 等. 池塘水产养殖精准投喂系统研制与试验[J]. 农业工程学报, 2023, 39(22): 27-34.
ZHAO Siqi, ZHAO Sanqin, GUO Zhenqi, et al. Development of the precise feeding system for pond culture[J]. Transactions of the Chinese Society of Agricultural Engineering (Transactions of the CSAE), 2023, 39(22): 27-34. (in Chinese with English abstract)
2026年第42卷第12期
PDF下载
232
85
引用本文
BibTeX
文章信息
doi: 10.11975/j.issn.1002-6819.202601127
  • 接收时间:2026-01-16
  • 首发时间:2026-08-20
  • 出版时间:2026-06-30
补充材料
相关文章
文章信息
作者
出版历史
  • 收稿日期:2026-01-16
  • 修回日期:2026-04-05
基金
作者信息
    1华中农业大学工学院,武汉 430070
    2农业农村部水产养殖设施工程重点实验室,武汉 430070

通讯作者:

谭鹤群,博士,教授,博士生导师,研究方向为畜禽水产养殖工程技术与装备。Email:
参考文献
分享链接
https://castjournals.cast.org.cn/joweb/nygcxb/CN/10.11975/j.issn.1002-6819.202601127
分享至
全文二维码

扫描看全文

引用本文
BibTeX
本文的引用情况
2种不同金属材料的力学参数

Family
属数
Number of
genus
种数
Number of
species
占总种数比例
Percentage of
total species (%)

Genus
种数
Number of
species
占总种数比例
Percentage of total
species (%)
鹅膏菌科Amanitaceae 2 11 5.26 鹅膏菌属 Amanita 10 4.78
小菇科 Mycenaceae 2 12 5.74 丝盖伞属 Inocybe 5 2.39
多孔菌科 Polyporaceae 8 14 6.70 蜡蘑属 Laccaria 5 2.39
红菇科 Russulaceae 3 23 11.00 小皮伞属 Marasmius 6 2.87
小菇属 Mycena 11 5.26
光柄菇属 Pluteus 5 2.39
红菇属 Russula 17 8.13
栓菌属 Trametes 5 2.39
关闭全屏