Article(id=1241116645033300311, tenantId=1146029695717560320, journalId=1234093305789726721, issueId=1241116641321350143, articleNumber=null, orderNo=null, doi=null, pmid=null, cstr=null, oa=null, hot=null, price=null, onlineType=0, articleFormat=0, articleType=null, articleTypeStr=null, receivedDate=1724083200000, receivedDateStr=2024-08-20, revisedDate=null, revisedDateStr=null, acceptedDate=null, acceptedDateStr=null, onlineDate=1773834867007, onlineDateStr=2026-03-18, pubDate=1742400000000, pubDateStr=2025-03-20, doiRegisterDate=null, doiRegisterDateStr=null, onlineIssueDate=1773834867007, onlineIssueDateStr=2026-03-18, onlineJustAcceptDate=null, onlineJustAcceptDateStr=null, onlineFirstDate=null, onlineFirstDateStr=null, sourceXml=null, magXml=null, createTime=1773834867007, creator=13701087609, updateTime=1773834867007, updator=13701087609, issue=Issue{id=1241116641321350143, tenantId=1146029695717560320, journalId=1234093305789726721, year='2025', volume='45', issue='3', pageStart='1185', pageEnd='1776', issueExtLink='null', onlineDate='null', pubDate='null', beforeIssueId=null, nextIssueId=null, price=null, status=1, issueComplete=1, articleOrder=1, issueType=-1, specialIssue=null, createTime=1773834866123, creator=13701087609, updateTime=1773881366030, updator=13701087609, preIssue=null, nextIssue=null, ext={EN=IssueExt(id=1241311676130193619, tenantId=1146029695717560320, journalId=1234093305789726721, issueId=1241116641321350143, language=EN, specialIssueTitle=, coverIllustrator=null, specialIssueEditor=, specialIssueAbout=), CN=IssueExt(id=1241311676130193620, tenantId=1146029695717560320, journalId=1234093305789726721, issueId=1241116641321350143, language=CN, specialIssueTitle=, coverIllustrator=null, specialIssueEditor=, specialIssueAbout=)}, issueFiles=null}, startPage=1231, endPage=1240, ext={EN=ArticleExt(id=1241116645394010461, articleId=1241116645033300311, tenantId=1146029695717560320, journalId=1234093305789726721, language=EN, title=Research on diesel vehicle NH3 emission prediction method based on CNN-Transformer fusion framework, columnId=1234106386020365051, journalTitle=China Environmental Science, columnName=Air Pollution Control, runingTitle=null, highlight=null, articleAbstract=

In this study, a diesel vehicle NH3 emission prediction model based on the fusion framework of Convolutional Neural Network(CNN)and Transformer is proposed. The model was developed by integrating the local feature extraction capability of CNN with the global dependency modeling capability of Transformer, enabling the highly accurate prediction of NH3 emissions from diesel vehicles under real road driving conditions. The study was conducted based on the actual on-road emissions test data of an N3-class diesel vehicle. Feature screening was performed using the Pearson correlation coefficient method, and the key hyperparameters of the model were optimized through the application of the Bayesian algorithm, which enhanced its performance. Additionally, the SHapley Additive exPlanations(SHAP)algorithm was utilized to identify the pivotal factors influencing NH3 emissions. The results indicated that the proposed model achieved highly accurate predictions of NH3 emissions from diesel vehicles in real road driving conditions when tested on an independent dataset. The R2, MAE, and MSE values of the predicted NH3 concentration compared to the actual measured values were 0.986, 0.663, and 2.285, respectively, which were significantly superior to those obtained by the traditional Random Forest(RF)model, the Long Short-Term Memory(LSTM)neural network model, and the Transformer model. This study provided an efficient and reliable method for monitoring NH3 emissions from in-use diesel vehicles and offered a novel perspective for elucidating the principal factors influencing NH3 emissions from diesel vehicles on the road.

, correspAuthors=Xiao-xin BAI, authorNote=null, correspAuthorsNote=null, copyrightStatement=null, 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, authorCompany=null, fund=null, authors=null, authorsList=Xiao-xin BAI, Xiang-yang GUO, Chun-ling WU, Feng-bin WANG, Xu LI, Wei-lin LIU), CN=ArticleExt(id=1241116658492821573, articleId=1241116645033300311, tenantId=1146029695717560320, journalId=1234093305789726721, language=CN, title=基于CNN-Transformer融合框架的柴油车氨排放预测方法, columnId=1234106388364981004, journalTitle=中国环境科学, columnName=大气污染与控制, runingTitle=null, highlight=null, articleAbstract=

本研究提出了一种基于卷积神经网络(CNN)和Transformer融合框架的柴油车NH3排放预测模型.该模型充分结合了CNN的局部特征提取能力和Transformer的全局依赖关注能力,实现了对实际行驶条件下柴油车NH3排放的高精度预测.研究以一辆N3类柴油车的实际道路排放测试数据为基础,采用Pearson相关系数法进行特征筛选,并利用贝叶斯优化算法对模型关键超参数进行调整,以提升模型性能.此外,应用SHAP算法量化了影响NH3排放的关键因素.结果表明,所提模型在独立数据集上能够高精度预测柴油车实际道路行驶中的NH3浓度排放,其预测的NH3浓度与实际测量值的R2、平均绝对误差(MAE)和均方误差(MSE)分别达到0.986、0.663和2.285,预测性能显著优于传统的随机森林(RF)模型、长短期记忆(LSTM)神经网络模型以及Transformer模型.研究为在用柴油车NH3排放监测提供了一种高效可靠的方法,同时为深入理解影响柴油车实际道路NH3排放的关键因素提供了新的研究思路.

, correspAuthors=白晓鑫, authorNote=null, correspAuthorsNote=
* 责任作者,工程师,
, copyrightStatement=null, copyrightOwner=null, extLink=null, articleAbsUrl=null, sourceXml=O9vqdTfn3NHlJKM3i9k4KQ==, magXml=CPH8k1//3pTVVxbH8Eikfw==, pdfUrl=null, pdf=Q8q5+L08TP7eY06KrDG+0w==, pdfFileSize=1808675, pdfExtLink=null, richHtmlUrl=null, mobilePdfUrl=null, reviewReport=null, pdfFirstPage=null, abstractGraph=eanG24aONmYrQXdyslb+mg==, abstractGraphContent=null, abstractVideo=null, citation=null, cebUrl=null, magXmlContent=S5M40a6uvr55lDejCJ1Sug==, mapNumber=null, authorCompany=null, fund=null, authors=

白晓鑫(1993-),男,山西阳泉人,工程师,硕士,主要从事重型发动机及整车排放控制研究.发表论文10余篇..

, authorsList=白晓鑫, 郭向阳, 吴春玲, 王凤滨, 李旭, 刘卫林)}, authors=[Author(id=1241116659054858385, tenantId=1146029695717560320, journalId=1234093305789726721, articleId=1241116645033300311, orderNo=0, firstName=null, middleName=null, lastName=null, nameCn=null, orcid=null, stid=null, country=null, authorPic=null, dead=0, email=baixiaoxin@catarc.ac.cn, emailSecond=null, emailThird=null, correspondingAuthor=1, authorType=1, ext={EN=AuthorExt(id=1241116659201659048, tenantId=1146029695717560320, journalId=1234093305789726721, articleId=1241116645033300311, authorId=1241116659054858385, language=EN, stringName=Xiao-xin BAI, firstName=Xiao-xin, middleName=null, lastName=BAI, prefix=null, suffix=null, authorComment=null, nameInitials=null, affiliation=null, department=null, xref=1, *, address=1.CATARC Automotive Test Center (Tianjin) Co., Ltd., Tianjin 300300, China, bio=null, bioImg=null, bioContent=null, aboutCorrespAuthor=null), CN=AuthorExt(id=1241116659302322361, tenantId=1146029695717560320, journalId=1234093305789726721, articleId=1241116645033300311, authorId=1241116659054858385, language=CN, stringName=白晓鑫, firstName=null, middleName=null, lastName=null, prefix=null, suffix=null, authorComment=null, nameInitials=null, affiliation=null, department=null, xref=1, *, address=1.中汽研汽车检验中心(天津)有限公司,天津 300300, bio={"content":"

白晓鑫(1993-),男,山西阳泉人,工程师,硕士,主要从事重型发动机及整车排放控制研究.发表论文10余篇..

"}, bioImg=null, bioContent=

白晓鑫(1993-),男,山西阳泉人,工程师,硕士,主要从事重型发动机及整车排放控制研究.发表论文10余篇..

, aboutCorrespAuthor=null)}, companyList=[AuthorCompany(id=1241116658765451365, tenantId=1146029695717560320, journalId=1234093305789726721, articleId=1241116645033300311, xref=1., ext=[AuthorCompanyExt(id=1241116658769645671, tenantId=1146029695717560320, journalId=1234093305789726721, articleId=1241116645033300311, companyId=1241116658765451365, language=EN, country=null, province=null, city=null, postcode=null, companyName=null, departmentName=null, remark=1.CATARC Automotive Test Center (Tianjin) Co., Ltd., Tianjin 300300, China), AuthorCompanyExt(id=1241116658782228586, tenantId=1146029695717560320, journalId=1234093305789726721, articleId=1241116645033300311, companyId=1241116658765451365, language=CN, country=null, province=null, city=null, postcode=null, companyName=null, departmentName=null, remark=1.中汽研汽车检验中心(天津)有限公司,天津 300300)])]), Author(id=1241116659411374285, tenantId=1146029695717560320, journalId=1234093305789726721, articleId=1241116645033300311, orderNo=1, 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=1241116659545592040, tenantId=1146029695717560320, journalId=1234093305789726721, articleId=1241116645033300311, authorId=1241116659411374285, language=EN, stringName=Xiang-yang GUO, firstName=Xiang-yang, middleName=null, lastName=GUO, prefix=null, suffix=null, authorComment=null, nameInitials=null, affiliation=null, department=null, xref=1, address=1.CATARC Automotive Test Center (Tianjin) Co., Ltd., Tianjin 300300, China, bio=null, bioImg=null, bioContent=null, aboutCorrespAuthor=null), CN=AuthorExt(id=1241116659650449656, tenantId=1146029695717560320, journalId=1234093305789726721, articleId=1241116645033300311, authorId=1241116659411374285, language=CN, stringName=郭向阳, firstName=null, middleName=null, lastName=null, prefix=null, suffix=null, authorComment=null, nameInitials=null, affiliation=null, department=null, xref=1, address=1.中汽研汽车检验中心(天津)有限公司,天津 300300, bio=null, bioImg=null, bioContent=null, aboutCorrespAuthor=null)}, companyList=[AuthorCompany(id=1241116658765451365, tenantId=1146029695717560320, journalId=1234093305789726721, articleId=1241116645033300311, xref=1., ext=[AuthorCompanyExt(id=1241116658769645671, tenantId=1146029695717560320, journalId=1234093305789726721, articleId=1241116645033300311, companyId=1241116658765451365, language=EN, country=null, province=null, city=null, postcode=null, companyName=null, departmentName=null, remark=1.CATARC Automotive Test Center (Tianjin) Co., Ltd., Tianjin 300300, China), AuthorCompanyExt(id=1241116658782228586, tenantId=1146029695717560320, journalId=1234093305789726721, articleId=1241116645033300311, companyId=1241116658765451365, language=CN, country=null, province=null, city=null, postcode=null, companyName=null, departmentName=null, remark=1.中汽研汽车检验中心(天津)有限公司,天津 300300)])]), Author(id=1241116659788861710, tenantId=1146029695717560320, journalId=1234093305789726721, articleId=1241116645033300311, 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=1241116659918885152, tenantId=1146029695717560320, journalId=1234093305789726721, articleId=1241116645033300311, authorId=1241116659788861710, language=EN, stringName=Chun-ling WU, firstName=Chun-ling, middleName=null, lastName=WU, prefix=null, suffix=null, authorComment=null, nameInitials=null, affiliation=null, department=null, xref=1, 2, address=1.CATARC Automotive Test Center (Tianjin) Co., Ltd., Tianjin 300300, China
2.School of Mechanical Engineering, Tianjin University, Tianjin 300072, China, bio=null, bioImg=null, bioContent=null, aboutCorrespAuthor=null), CN=AuthorExt(id=1241116660044714288, tenantId=1146029695717560320, journalId=1234093305789726721, articleId=1241116645033300311, authorId=1241116659788861710, language=CN, stringName=吴春玲, firstName=null, middleName=null, lastName=null, prefix=null, suffix=null, authorComment=null, nameInitials=null, affiliation=null, department=null, xref=1, 2, address=1.中汽研汽车检验中心(天津)有限公司,天津 300300
2.天津大学机械工程学院,天津 300072, bio=null, bioImg=null, bioContent=null, aboutCorrespAuthor=null)}, companyList=[AuthorCompany(id=1241116658765451365, tenantId=1146029695717560320, journalId=1234093305789726721, articleId=1241116645033300311, xref=1., ext=[AuthorCompanyExt(id=1241116658769645671, tenantId=1146029695717560320, journalId=1234093305789726721, articleId=1241116645033300311, companyId=1241116658765451365, language=EN, country=null, province=null, city=null, postcode=null, companyName=null, departmentName=null, remark=1.CATARC Automotive Test Center (Tianjin) Co., Ltd., Tianjin 300300, China), AuthorCompanyExt(id=1241116658782228586, tenantId=1146029695717560320, journalId=1234093305789726721, articleId=1241116645033300311, companyId=1241116658765451365, language=CN, country=null, province=null, city=null, postcode=null, companyName=null, departmentName=null, remark=1.中汽研汽车检验中心(天津)有限公司,天津 300300)]), AuthorCompany(id=1241116658878697590, tenantId=1146029695717560320, journalId=1234093305789726721, articleId=1241116645033300311, xref=2., ext=[AuthorCompanyExt(id=1241116658891280507, tenantId=1146029695717560320, journalId=1234093305789726721, articleId=1241116645033300311, companyId=1241116658878697590, language=EN, country=null, province=null, city=null, postcode=null, companyName=null, departmentName=null, remark=2.School of Mechanical Engineering, Tianjin University, Tianjin 300072, China), AuthorCompanyExt(id=1241116658916446333, tenantId=1146029695717560320, journalId=1234093305789726721, articleId=1241116645033300311, companyId=1241116658878697590, language=CN, country=null, province=null, city=null, postcode=null, companyName=null, departmentName=null, remark=2.天津大学机械工程学院,天津 300072)])]), Author(id=1241116660183126337, tenantId=1146029695717560320, journalId=1234093305789726721, articleId=1241116645033300311, 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=1241116660296372566, tenantId=1146029695717560320, journalId=1234093305789726721, articleId=1241116645033300311, authorId=1241116660183126337, language=EN, stringName=Feng-bin WANG, firstName=Feng-bin, middleName=null, lastName=WANG, prefix=null, suffix=null, authorComment=null, nameInitials=null, affiliation=null, department=null, xref=1, address=1.CATARC Automotive Test Center (Tianjin) Co., Ltd., Tianjin 300300, China, bio=null, bioImg=null, bioContent=null, aboutCorrespAuthor=null), CN=AuthorExt(id=1241116660384452961, tenantId=1146029695717560320, journalId=1234093305789726721, articleId=1241116645033300311, authorId=1241116660183126337, language=CN, stringName=王凤滨, firstName=null, middleName=null, lastName=null, prefix=null, suffix=null, authorComment=null, nameInitials=null, affiliation=null, department=null, xref=1, address=1.中汽研汽车检验中心(天津)有限公司,天津 300300, bio=null, bioImg=null, bioContent=null, aboutCorrespAuthor=null)}, companyList=[AuthorCompany(id=1241116658765451365, tenantId=1146029695717560320, journalId=1234093305789726721, articleId=1241116645033300311, xref=1., ext=[AuthorCompanyExt(id=1241116658769645671, tenantId=1146029695717560320, journalId=1234093305789726721, articleId=1241116645033300311, companyId=1241116658765451365, language=EN, country=null, province=null, city=null, postcode=null, companyName=null, departmentName=null, remark=1.CATARC Automotive Test Center (Tianjin) Co., Ltd., Tianjin 300300, China), AuthorCompanyExt(id=1241116658782228586, tenantId=1146029695717560320, journalId=1234093305789726721, articleId=1241116645033300311, companyId=1241116658765451365, language=CN, country=null, province=null, city=null, postcode=null, companyName=null, departmentName=null, remark=1.中汽研汽车检验中心(天津)有限公司,天津 300300)])]), Author(id=1241116660472533363, tenantId=1146029695717560320, journalId=1234093305789726721, articleId=1241116645033300311, orderNo=4, 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=1241116660585779586, tenantId=1146029695717560320, journalId=1234093305789726721, articleId=1241116645033300311, authorId=1241116660472533363, language=EN, stringName=Xu LI, firstName=Xu, middleName=null, lastName=LI, prefix=null, suffix=null, authorComment=null, nameInitials=null, affiliation=null, department=null, xref=1, address=1.CATARC Automotive Test Center (Tianjin) Co., Ltd., Tianjin 300300, China, bio=null, bioImg=null, bioContent=null, aboutCorrespAuthor=null), CN=AuthorExt(id=1241116660682248591, tenantId=1146029695717560320, journalId=1234093305789726721, articleId=1241116645033300311, authorId=1241116660472533363, language=CN, stringName=李旭, firstName=null, middleName=null, lastName=null, prefix=null, suffix=null, authorComment=null, nameInitials=null, affiliation=null, department=null, xref=1, address=1.中汽研汽车检验中心(天津)有限公司,天津 300300, bio=null, bioImg=null, bioContent=null, aboutCorrespAuthor=null)}, companyList=[AuthorCompany(id=1241116658765451365, tenantId=1146029695717560320, journalId=1234093305789726721, articleId=1241116645033300311, xref=1., ext=[AuthorCompanyExt(id=1241116658769645671, tenantId=1146029695717560320, journalId=1234093305789726721, articleId=1241116645033300311, companyId=1241116658765451365, language=EN, country=null, province=null, city=null, postcode=null, companyName=null, departmentName=null, remark=1.CATARC Automotive Test Center (Tianjin) Co., Ltd., Tianjin 300300, China), AuthorCompanyExt(id=1241116658782228586, tenantId=1146029695717560320, journalId=1234093305789726721, articleId=1241116645033300311, companyId=1241116658765451365, language=CN, country=null, province=null, city=null, postcode=null, companyName=null, departmentName=null, remark=1.中汽研汽车检验中心(天津)有限公司,天津 300300)])]), Author(id=1241116660782911897, tenantId=1146029695717560320, journalId=1234093305789726721, articleId=1241116645033300311, orderNo=5, 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=1241116660879380905, tenantId=1146029695717560320, journalId=1234093305789726721, articleId=1241116645033300311, authorId=1241116660782911897, language=EN, stringName=Wei-lin LIU, firstName=Wei-lin, middleName=null, lastName=LIU, prefix=null, suffix=null, authorComment=null, nameInitials=null, affiliation=null, department=null, xref=1, address=1.CATARC Automotive Test Center (Tianjin) Co., Ltd., Tianjin 300300, China, bio=null, bioImg=null, bioContent=null, aboutCorrespAuthor=null), CN=AuthorExt(id=1241116660996821433, tenantId=1146029695717560320, journalId=1234093305789726721, articleId=1241116645033300311, authorId=1241116660782911897, language=CN, stringName=刘卫林, firstName=null, middleName=null, lastName=null, prefix=null, suffix=null, authorComment=null, nameInitials=null, affiliation=null, department=null, xref=1, address=1.中汽研汽车检验中心(天津)有限公司,天津 300300, bio=null, bioImg=null, bioContent=null, aboutCorrespAuthor=null)}, companyList=[AuthorCompany(id=1241116658765451365, tenantId=1146029695717560320, journalId=1234093305789726721, articleId=1241116645033300311, xref=1., ext=[AuthorCompanyExt(id=1241116658769645671, tenantId=1146029695717560320, journalId=1234093305789726721, articleId=1241116645033300311, companyId=1241116658765451365, language=EN, country=null, province=null, city=null, postcode=null, companyName=null, departmentName=null, remark=1.CATARC Automotive Test Center (Tianjin) Co., Ltd., Tianjin 300300, China), AuthorCompanyExt(id=1241116658782228586, tenantId=1146029695717560320, journalId=1234093305789726721, articleId=1241116645033300311, companyId=1241116658765451365, language=CN, country=null, province=null, city=null, postcode=null, companyName=null, departmentName=null, remark=1.中汽研汽车检验中心(天津)有限公司,天津 300300)])])], keywords=[Keyword(id=1241116661164593609, tenantId=1146029695717560320, journalId=1234093305789726721, articleId=1241116645033300311, language=EN, orderNo=1, keyword=diesel vehicles), Keyword(id=1241116661361725912, tenantId=1146029695717560320, journalId=1234093305789726721, articleId=1241116645033300311, language=EN, orderNo=2, keyword=emission), Keyword(id=1241116662754234852, tenantId=1146029695717560320, journalId=1234093305789726721, articleId=1241116645033300311, language=EN, orderNo=3, keyword=NH3), Keyword(id=1241116662959755762, tenantId=1146029695717560320, journalId=1234093305789726721, articleId=1241116645033300311, language=EN, orderNo=4, keyword=convolutional neural network), Keyword(id=1241116663110750728, tenantId=1146029695717560320, journalId=1234093305789726721, articleId=1241116645033300311, language=EN, orderNo=5, keyword=transformer), Keyword(id=1241116663228191254, tenantId=1146029695717560320, journalId=1234093305789726721, articleId=1241116645033300311, language=CN, orderNo=1, keyword=柴油车), Keyword(id=1241116663387574825, tenantId=1146029695717560320, journalId=1234093305789726721, articleId=1241116645033300311, language=CN, orderNo=2, keyword=排放), Keyword(id=1241116663559541306, tenantId=1146029695717560320, journalId=1234093305789726721, articleId=1241116645033300311, language=CN, orderNo=3, keyword=NH3), Keyword(id=1241116663681176136, tenantId=1146029695717560320, journalId=1234093305789726721, articleId=1241116645033300311, language=CN, orderNo=4, keyword=卷积神经网络), Keyword(id=1241116663836365405, tenantId=1146029695717560320, journalId=1234093305789726721, articleId=1241116645033300311, language=CN, orderNo=5, keyword=Transformer)], refs=[Reference(id=1241116669460926523, tenantId=1146029695717560320, journalId=1234093305789726721, articleId=1241116645033300311, doi=null, pmid=null, pmcid=null, year=2024, volume=52, issue=2, pageStart=48, pageEnd=62, url=null, language=null, rfNumber=[1], rfOrder=0, authorNames=null, journalName=环境保护, refType=null, unstructuredReference=中国移动源环境管理年报(2023年) [J]. 环境保护202452(2):48-62., articleTitle=中国移动源环境管理年报(2023年), refAbstract=null), Reference(id=1241116669578367047, tenantId=1146029695717560320, journalId=1234093305789726721, articleId=1241116645033300311, doi=null, pmid=null, pmcid=null, year=2024, volume=52, issue=2, pageStart=48, pageEnd=62, url=null, language=null, rfNumber=[1], rfOrder=1, authorNames=null, journalName=Environmental Protection, refType=null, unstructuredReference=China mobile source environmental management annual report (2023)[J]. Environmental Protection202452(2):48-62., articleTitle=China mobile source environmental management annual report (2023), refAbstract=null), Reference(id=1241116669674836048, tenantId=1146029695717560320, journalId=1234093305789726721, articleId=1241116645033300311, doi=null, pmid=null, pmcid=null, year=2003, volume=29, issue=2/3, pageStart=277, pageEnd=286, url=null, language=null, rfNumber=[2], rfOrder=2, authorNames=Anderson N, Strader R, Davidson C, journalName=Environment international, refType=null, unstructuredReference=Anderson NStrader RDavidson C. Airborne reduced nitrogen: Ammonia emissions from agriculture and other sources [J]. Environment international200329(2/3): 277-286., articleTitle=Airborne reduced nitrogen: Ammonia emissions from agriculture and other sources, refAbstract=null), Reference(id=1241116669809053788, tenantId=1146029695717560320, journalId=1234093305789726721, articleId=1241116645033300311, doi=null, pmid=null, pmcid=null, year=2010, volume=44, issue=29, pageStart=3547, pageEnd=3557, url=null, language=null, rfNumber=[3], rfOrder=3, authorNames=Zhang H, Ying Q, journalName=Atmospheric Environment, refType=null, unstructuredReference=Zhang HYing Q. Source apportionment of airborne particulate matter in Southeast Texas using a source-oriented 3D air quality model [J]. Atmospheric Environment201044(29):3547-3557., articleTitle=Source apportionment of airborne particulate matter in Southeast Texas using a source-oriented 3D air quality model, refAbstract=null), Reference(id=1241116669947465829, tenantId=1146029695717560320, journalId=1234093305789726721, articleId=1241116645033300311, doi=null, pmid=null, pmcid=null, year=2020, volume=54, issue=24, pageStart=15689, pageEnd=15697, url=null, language=null, rfNumber=[4], rfOrder=4, authorNames=Farren, N J., Davison J, Rose R A, journalName=Environmental science & technology, refType=null, unstructuredReference=Farren, N J.Davison JRose R A,et al. Underestimated ammonia emissions from road vehicles [J]. Environmental science & technology202054(24):15689-15697., articleTitle=Underestimated ammonia emissions from road vehicles, refAbstract=null), Reference(id=1241116670136209524, tenantId=1146029695717560320, journalId=1234093305789726721, articleId=1241116645033300311, doi=null, pmid=null, pmcid=null, year=2023, volume=43, issue=6, pageStart=2673, pageEnd=2682, url=null, language=null, rfNumber=[5], rfOrder=5, authorNames=陈培林, 肖欣欣, 王勤耕, journalName=中国环境科学, refType=null, unstructuredReference=陈培林,肖欣欣,王勤耕. 基于卫星观测的2010~2020年中国高分辨率NH3排放特征 [J]. 中国环境科学202343(6):2673-2682., articleTitle=基于卫星观测的2010~2020年中国高分辨率NH3排放特征, refAbstract=null), Reference(id=1241116670274621564, tenantId=1146029695717560320, journalId=1234093305789726721, articleId=1241116645033300311, doi=null, pmid=null, pmcid=null, year=2023, volume=43, issue=6, pageStart=2673, pageEnd=2682, url=null, language=null, rfNumber=[5], rfOrder=6, authorNames=Chen P L, Xiao X X, Wang Q G, journalName=China Environmental Science, refType=null, unstructuredReference=Chen P LXiao X XWang Q G,et al. High-resolution characteristics of NH3 emission from 2010 to 2020 in China based on satellite observation [J]. China Environmental Science202343(6):2673-2682., articleTitle=High-resolution characteristics of NH3 emission from 2010 to 2020 in China based on satellite observation, refAbstract=null), Reference(id=1241116671797153926, tenantId=1146029695717560320, journalId=1234093305789726721, articleId=1241116645033300311, doi=null, pmid=null, pmcid=null, year=2024, volume=49, issue=12, pageStart=89, pageEnd=94, url=null, language=null, rfNumber=[6], rfOrder=7, authorNames=白晓鑫, 吴春玲, 刘卫林, journalName=汽车实用技术, refType=null, unstructuredReference=白晓鑫,吴春玲,刘卫林,等. 柴油车尿素溶液品质在线检测方法研究. 汽车实用技术202449(12):89-94., articleTitle=柴油车尿素溶液品质在线检测方法研究, refAbstract=null), Reference(id=1241116671960731795, tenantId=1146029695717560320, journalId=1234093305789726721, articleId=1241116645033300311, doi=null, pmid=null, pmcid=null, year=2024, volume=49, issue=12, pageStart=89, pageEnd=94, url=null, language=null, rfNumber=[6], rfOrder=8, authorNames=Bai X X, Wu C L, Liu W L, journalName=Automobile Applied Technology, refType=null, unstructuredReference=Bai X XWu C LLiu W L,et al. Research on online detection method for urea solution quality of diesel vehicle [J]. Automobile Applied Technology202449(12):89-94., articleTitle=Research on online detection method for urea solution quality of diesel vehicle, refAbstract=null), Reference(id=1241116672103338142, tenantId=1146029695717560320, journalId=1234093305789726721, articleId=1241116645033300311, doi=null, pmid=null, pmcid=null, year=null, volume=null, issue=null, pageStart=null, pageEnd=null, url=null, language=null, rfNumber=[7], rfOrder=9, authorNames=null, journalName=null, refType=null, unstructuredReference=GB 17691—2018 重型柴油车污染物排放限值及测量方法(中国第六阶段) [S]., articleTitle=null, refAbstract=null), Reference(id=1241116672296276140, tenantId=1146029695717560320, journalId=1234093305789726721, articleId=1241116645033300311, doi=null, pmid=null, pmcid=null, year=null, volume=null, issue=null, pageStart=null, pageEnd=null, url=null, language=null, rfNumber=[7], rfOrder=10, authorNames=null, journalName=null, refType=null, unstructuredReference=GB 17697-2018 Limits and measurement methods for emissions from diesel-fueled heavy-duty vehicles (China VI) [S]., articleTitle=null, refAbstract=null), Reference(id=1241116672413716662, tenantId=1146029695717560320, journalId=1234093305789726721, articleId=1241116645033300311, doi=null, pmid=null, pmcid=null, year=2017, volume=166, issue=null, pageStart=488, pageEnd=497, url=null, language=null, rfNumber=[8], rfOrder=11, authorNames=Suarez-Bertoa R, Mendoza-Villafuerte P, Riccobono F, journalName=Atmospheric Environment, refType=null, unstructuredReference=Suarez-Bertoa RMendoza-Villafuerte PRiccobono F,et al. On-road measurement of NH3 emissions from gasoline and diesel passenger cars during real world driving conditions [J]. Atmospheric Environment2017166:488-497., articleTitle=On-road measurement of NH3 emissions from gasoline and diesel passenger cars during real world driving conditions, refAbstract=null), Reference(id=1241116672547934396, tenantId=1146029695717560320, journalId=1234093305789726721, articleId=1241116645033300311, doi=null, pmid=null, pmcid=null, year=2024, volume=316, issue=null, pageStart=null, pageEnd=null, url=null, language=null, rfNumber=[9], rfOrder=12, authorNames=Zhu H, Ma T, Toumasatos Z, journalName=Atmospheric Environment, refType=null, unstructuredReference=Zhu HMa TToumasatos Z,et al. On-road NOx and NH3emissions measurements from in-use heavy-duty diesel and natural gas trucks in the South Coast air Basin of California [J]. Atmospheric Environment2024316:120179., articleTitle=On-road NOx and NH3emissions measurements from in-use heavy-duty diesel and natural gas trucks in the South Coast air Basin of California, refAbstract=null), Reference(id=1241116672707317966, tenantId=1146029695717560320, journalId=1234093305789726721, articleId=1241116645033300311, doi=null, pmid=null, pmcid=null, year=2017, volume=609, issue=null, pageStart=546, pageEnd=555, url=null, language=null, rfNumber=[10], rfOrder=13, authorNames=Mendoza-Villafuerte P, Suarez-Bertoa R, Giechaskiel B, journalName=Science of the Total Environment, refType=null, unstructuredReference=Mendoza-Villafuerte PSuarez-Bertoa RGiechaskiel B,et al. NOx,NH3,N2O and PN real driving emissions from a Euro VI heavy-duty vehicle. Impact of regulatory on-road test conditions on emissions [J]. Science of the Total Environment2017609:546-555., articleTitle=NOx,NH3,N2O and PN real driving emissions from a Euro VI heavy-duty vehicle. Impact of regulatory on-road test conditions on emissions, refAbstract=null), Reference(id=1241116672824758489, tenantId=1146029695717560320, journalId=1234093305789726721, articleId=1241116645033300311, doi=null, pmid=null, pmcid=null, year=null, volume=null, issue=null, pageStart=null, pageEnd=null, url=null, language=null, rfNumber=[11], rfOrder=14, authorNames=European Commission, journalName=null, refType=null, unstructuredReference=European Commission. Proposal for a regulation of the European parliament and of the council on type-approval of motor vehicles and of engines and of systems,components and separate technical units intended for such vehicles,with respect to their emissions and battery durability (Euro 7) [EB/OL]. (2022-11-10)[2024-06-04]. https://eur-lex.europa.eu/legal-content/EN/TXT/?uri=CELEX%3A52022PC0586., articleTitle=Proposal for a regulation of the European parliament and of the council on type-approval of motor vehicles and of engines and of systems,components and separate technical units intended for such vehicles,with respect to their emissions and battery durability (Euro 7), refAbstract=null), Reference(id=1241116672979947751, tenantId=1146029695717560320, journalId=1234093305789726721, articleId=1241116645033300311, doi=null, pmid=null, pmcid=null, year=2013, volume=64, issue=null, pageStart=329, pageEnd=338, url=null, language=null, rfNumber=[12], rfOrder=15, authorNames=Kousoulidou M, Georgios F, Leonidas N, journalName=Atmospheric Environment, refType=null, unstructuredReference=Kousoulidou MGeorgios FLeonidas N,et al. Use of portable emissions measurement system (PEMS) for the development and validation of passenger car emission factors [J]. Atmospheric Environment201364:329-338., articleTitle=Use of portable emissions measurement system (PEMS) for the development and validation of passenger car emission factors, refAbstract=null), Reference(id=1241116673143525623, tenantId=1146029695717560320, journalId=1234093305789726721, articleId=1241116645033300311, doi=null, pmid=null, pmcid=null, year=2015, volume=49, issue=8, pageStart=5236, pageEnd=5244, url=null, language=null, rfNumber=[13], rfOrder=16, authorNames=Thiruvengadam A, Besch M C, Thiruvengadam P, journalName=Environmental Science & Technology, refType=null, unstructuredReference=Thiruvengadam ABesch M CThiruvengadam P,et al. Emission rates of regulated pollutants from current technology heavy-duty diesel and natural gas goods movement vehicles [J]. Environmental Science & Technology201549(8):5236-5244., articleTitle=Emission rates of regulated pollutants from current technology heavy-duty diesel and natural gas goods movement vehicles, refAbstract=null), Reference(id=1241116673286131969, tenantId=1146029695717560320, journalId=1234093305789726721, articleId=1241116645033300311, doi=null, pmid=null, pmcid=null, year=2018, volume=52, issue=19, pageStart=11223, pageEnd=11231, url=null, language=null, rfNumber=[14], rfOrder=17, authorNames=Huang C, Hu Q, Lou S, journalName=Environmental Science & Technology, refType=null, unstructuredReference=Huang CHu QLou S,et al. Ammonia emission measurements for light-duty gasoline vehicles in China and implications for emission modeling [J]. Environmental Science & Technology201852(19):11223-11231., articleTitle=Ammonia emission measurements for light-duty gasoline vehicles in China and implications for emission modeling, refAbstract=null), Reference(id=1241116673579733262, tenantId=1146029695717560320, journalId=1234093305789726721, articleId=1241116645033300311, doi=null, pmid=null, pmcid=null, year=2021, volume=22, issue=10, pageStart=3209, pageEnd=3218, url=null, language=null, rfNumber=[15], rfOrder=18, authorNames=Pla B, Piqueras P, Bares P, journalName=International Journal of Engine Research, refType=null, unstructuredReference=Pla BPiqueras PBares P,et al. NOx sensor cross sensitivity model and simultaneous prediction of NOx and NH3 slip from automotive catalytic converters under real driving conditions [J]. International Journal of Engine Research202122(10):3209-3218., articleTitle=NOx sensor cross sensitivity model and simultaneous prediction of NOx and NH3 slip from automotive catalytic converters under real driving conditions, refAbstract=null), Reference(id=1241116673726533911, tenantId=1146029695717560320, journalId=1234093305789726721, articleId=1241116645033300311, doi=null, pmid=null, pmcid=null, year=2017, volume=65, issue=7, pageStart=5990, pageEnd=5998, url=null, language=null, rfNumber=[16], rfOrder=19, authorNames=Wen L, Li X, Gao L, journalName=IEEE Transactions on Industrial Electronics, refType=null, unstructuredReference=Wen LLi XGao L,et al. A new convolutional neural network-based data-driven fault diagnosis method [J]. IEEE Transactions on Industrial Electronics201765(7):5990-5998., articleTitle=A new convolutional neural network-based data-driven fault diagnosis method, refAbstract=null), Reference(id=1241116673864945955, tenantId=1146029695717560320, journalId=1234093305789726721, articleId=1241116645033300311, doi=null, pmid=null, pmcid=null, year=2017, volume=null, issue=null, pageStart=5998, pageEnd=6008, url=null, language=null, rfNumber=[17], rfOrder=20, authorNames=Vaswani A, Shazeer N, Parmar N, journalName=Advances in Neural Information Processing Systems, refType=null, unstructuredReference=Vaswani AShazeer NParmar N,et al. Attention is all you need [J]. Advances in Neural Information Processing Systems2017:5998-6008., articleTitle=Attention is all you need, refAbstract=null), Reference(id=1241116673986580782, tenantId=1146029695717560320, journalId=1234093305789726721, articleId=1241116645033300311, doi=null, pmid=null, pmcid=null, year=2024, volume=27, issue=6, pageStart=166, pageEnd=171, url=null, language=null, rfNumber=[18], rfOrder=21, authorNames=黄茂庭, 徐金明, journalName=城市轨道交通研究, refType=null, unstructuredReference=黄茂庭,徐金明. 使用CNN(卷积神经网络)-LSTM(长短期记忆)联合神经网络预测盾构隧道施工引起的地面沉降 [J]. 城市轨道交通研究202427(6):166-171., articleTitle=使用CNN(卷积神经网络)-LSTM(长短期记忆)联合神经网络预测盾构隧道施工引起的地面沉降, refAbstract=null), Reference(id=1241116674095632696, tenantId=1146029695717560320, journalId=1234093305789726721, articleId=1241116645033300311, doi=null, pmid=null, pmcid=null, year=2024, volume=27, issue=6, pageStart=166, pageEnd=171, url=null, language=null, rfNumber=[18], rfOrder=22, authorNames=Huang M T, Xu J M, journalName=Urban Mass Transit, refType=null, unstructuredReference=Huang M TXu J M. Prediction of land subsidence caused by shield tunnel construction with joint CNN-LSTM neural network [J]. Urban Mass Transit202427(6):166-171., articleTitle=Prediction of land subsidence caused by shield tunnel construction with joint CNN-LSTM neural network, refAbstract=null), Reference(id=1241116674242433346, tenantId=1146029695717560320, journalId=1234093305789726721, articleId=1241116645033300311, doi=null, pmid=null, pmcid=null, year=2018, volume=null, issue=null, pageStart=1, pageEnd=5, url=null, language=null, rfNumber=[19], rfOrder=23, authorNames=Thakkar V, Tewary S, Chakraborty C, journalName=null, refType=null, unstructuredReference=Thakkar VTewary SChakraborty C. Batch Normalization in convolutional neural networks—A comparative study with CIFAR-10 data [C]//2018fifth international conference on emerging applications of information technology (EAIT). IEEE,2018:1-5., articleTitle=Batch Normalization in convolutional neural networks—A comparative study with CIFAR-10 data, refAbstract=null), Reference(id=1241116674368262475, tenantId=1146029695717560320, journalId=1234093305789726721, articleId=1241116645033300311, doi=null, pmid=null, pmcid=null, year=2017, volume=30, issue=null, pageStart=5998, pageEnd=6008, url=null, language=null, rfNumber=[20], rfOrder=24, authorNames=Vaswani A, Shazeer N, Parmar N, journalName=Advances in Neural Information Processing Systems, refType=null, unstructuredReference=Vaswani AShazeer NParmar N,et al. Attention is all you need [J]. Advances in Neural Information Processing Systems201730:5998-6008., articleTitle=Attention is all you need, refAbstract=null), Reference(id=1241116674473120080, tenantId=1146029695717560320, journalId=1234093305789726721, articleId=1241116645033300311, doi=null, pmid=null, pmcid=null, year=1989, volume=110, issue=11, pageStart=916, pageEnd=921, url=null, language=null, rfNumber=[21], rfOrder=25, authorNames=Williamson D F, Parker R A, Kendrick J S, journalName=Annals of internal medicine, refType=null, unstructuredReference=Williamson D FParker R AKendrick J S. The box plot: a simple visual method to interpret data [J]. Annals of internal medicine1989110(11):916-921., articleTitle=The box plot: a simple visual method to interpret data, refAbstract=null), Reference(id=1241116674594754909, tenantId=1146029695717560320, journalId=1234093305789726721, articleId=1241116645033300311, doi=null, pmid=null, pmcid=null, year=2020, volume=null, issue=null, pageStart=null, pageEnd=null, url=null, language=null, rfNumber=[22], rfOrder=26, authorNames=陈婉娇, journalName=null, refType=null, unstructuredReference=陈婉娇. 缺失数据插补方法及其在医学领域的应用研究 [D]. 广州:华南理工大学,2020., articleTitle=缺失数据插补方法及其在医学领域的应用研究, refAbstract=null), Reference(id=1241116674716389737, tenantId=1146029695717560320, journalId=1234093305789726721, articleId=1241116645033300311, doi=null, pmid=null, pmcid=null, year=2020, volume=null, issue=null, pageStart=null, pageEnd=null, url=null, language=null, rfNumber=[22], rfOrder=27, authorNames=Chen W J, journalName=null, refType=null, unstructuredReference=Chen W J. Research on application of missing data lmputation in medical field [D]. Guang Zhou: South China University of Technology,2020., articleTitle=Research on application of missing data lmputation in medical field, refAbstract=null), Reference(id=1241116676234727790, tenantId=1146029695717560320, journalId=1234093305789726721, articleId=1241116645033300311, doi=null, pmid=null, pmcid=null, year=2024, volume=465, issue=null, pageStart=null, pageEnd=null, url=null, language=null, rfNumber=[23], rfOrder=28, authorNames=He L, Li G, Wu X, journalName=Journal of Hazardous Materials, refType=null, unstructuredReference=He LLi GWu X,et al. Characteristics of NOx and NH3 emissions from in-use heavy-duty diesel vehicles with various aftertreatment technologies in China [J]. Journal of Hazardous Materials2024465:133073., articleTitle=Characteristics of NOx and NH3 emissions from in-use heavy-duty diesel vehicles with various aftertreatment technologies in China, refAbstract=null), Reference(id=1241116676503163259, tenantId=1146029695717560320, journalId=1234093305789726721, articleId=1241116645033300311, doi=null, pmid=null, pmcid=null, year=2022, volume=null, issue=null, pageStart=null, pageEnd=null, url=null, language=null, rfNumber=[24], rfOrder=29, authorNames=Chen X, Liang C, Huang D, journalName=null, refType=null, unstructuredReference=Chen XLiang CHuang D,et al. Evolved optimizer for vision[C]//First Conference on Automated Machine Learning (Late-Breaking Workshop). 2022., articleTitle=Evolved optimizer for vision, refAbstract=null), Reference(id=1241116676616409477, tenantId=1146029695717560320, journalId=1234093305789726721, articleId=1241116645033300311, doi=10.48550/arXiv.1608.03983, pmid=null, pmcid=null, year=2016, volume=null, issue=null, pageStart=null, pageEnd=null, url=null, language=null, rfNumber=[25], rfOrder=30, authorNames=Loshchilov I, Hutter F, journalName=null, refType=null, unstructuredReference=Loshchilov IHutter F. SGDR: Stochastic gradient descent with warm restarts [J]. 2016., articleTitle=SGDR: Stochastic gradient descent with warm restarts, refAbstract=null), Reference(id=1241116676742238606, tenantId=1146029695717560320, journalId=1234093305789726721, articleId=1241116645033300311, doi=null, pmid=null, pmcid=null, year=2023, volume=48, issue=5, pageStart=56, pageEnd=63, url=null, language=null, rfNumber=[26], rfOrder=31, authorNames=白晓鑫, 吴春玲, 景晓军, journalName=汽车实用技术, refType=null, unstructuredReference=白晓鑫,吴春玲,景晓军,等. 基于RLS和BO算法的重型车载重估算研究 [J]. 汽车实用技术202348(5):56-63., articleTitle=基于RLS和BO算法的重型车载重估算研究, refAbstract=null), Reference(id=1241116676905816471, tenantId=1146029695717560320, journalId=1234093305789726721, articleId=1241116645033300311, doi=null, pmid=null, pmcid=null, year=2023, volume=48, issue=5, pageStart=56, pageEnd=63, url=null, language=null, rfNumber=[26], rfOrder=32, authorNames=Bai X X, Wu C L, Jing X J, journalName=Automobile Applied Technology, refType=null, unstructuredReference=Bai X XWu C LJing X J,et al. Research on heavy-duty vehicle mass estimation based on recursive least square and bayesian optimization algorithm [J]. Automobile Applied Technology202348(5):56-63., articleTitle=Research on heavy-duty vehicle mass estimation based on recursive least square and bayesian optimization algorithm, refAbstract=null), Reference(id=1241116677019062689, tenantId=1146029695717560320, journalId=1234093305789726721, articleId=1241116645033300311, doi=null, pmid=null, pmcid=null, year=2018, volume=30, issue=1, pageStart=197, pageEnd=215, url=null, language=null, rfNumber=[27], rfOrder=33, authorNames=Jung Y, journalName=Journal of nonparametric statistics, refType=null, unstructuredReference=Jung Y. Multiple predicting K-fold cross-validation for model selection [J]. Journal of nonparametric statistics201830(1):197-215., articleTitle=Multiple predicting K-fold cross-validation for model selection, refAbstract=null), Reference(id=1241116677123920298, tenantId=1146029695717560320, journalId=1234093305789726721, articleId=1241116645033300311, doi=null, pmid=null, pmcid=null, year=2020, volume=null, issue=null, pageStart=340, pageEnd=347, url=null, language=null, rfNumber=[28], rfOrder=34, authorNames=Marcílio W E, Eler D M, journalName=null, refType=null, unstructuredReference=Marcílio W EEler D M. From explanations to feature selection: assessing SHAP values as feature selection mechanism[C]//2020 33rd SIBGRAPI conference on Graphics,Patterns and Images (SIBGRAPI).IEEE,2020: 340-347., articleTitle=From explanations to feature selection: assessing SHAP values as feature selection mechanism, refAbstract=null)], funds=[Fund(id=1241116669326708781, tenantId=1146029695717560320, journalId=1234093305789726721, articleId=1241116645033300311, awardId=2022YFC3701800, language=CN, fundingSource=国家重点研发计划(2022YFC3701800), fundOrder=null, country=null)], companyList=[AuthorCompany(id=1241116658765451365, tenantId=1146029695717560320, journalId=1234093305789726721, articleId=1241116645033300311, xref=1., ext=[AuthorCompanyExt(id=1241116658769645671, tenantId=1146029695717560320, journalId=1234093305789726721, articleId=1241116645033300311, companyId=1241116658765451365, language=EN, country=null, province=null, city=null, postcode=null, companyName=null, departmentName=null, remark=1.CATARC Automotive Test Center (Tianjin) Co., Ltd., Tianjin 300300, China), AuthorCompanyExt(id=1241116658782228586, tenantId=1146029695717560320, journalId=1234093305789726721, articleId=1241116645033300311, companyId=1241116658765451365, language=CN, country=null, province=null, city=null, postcode=null, companyName=null, departmentName=null, remark=1.中汽研汽车检验中心(天津)有限公司,天津 300300)]), AuthorCompany(id=1241116658878697590, tenantId=1146029695717560320, journalId=1234093305789726721, articleId=1241116645033300311, xref=2., ext=[AuthorCompanyExt(id=1241116658891280507, tenantId=1146029695717560320, journalId=1234093305789726721, articleId=1241116645033300311, companyId=1241116658878697590, language=EN, country=null, province=null, city=null, postcode=null, companyName=null, departmentName=null, remark=2.School of Mechanical Engineering, Tianjin University, Tianjin 300072, China), AuthorCompanyExt(id=1241116658916446333, tenantId=1146029695717560320, journalId=1234093305789726721, articleId=1241116645033300311, companyId=1241116658878697590, language=CN, country=null, province=null, city=null, postcode=null, companyName=null, departmentName=null, remark=2.天津大学机械工程学院,天津 300072)])], figs=[ArticleFig(id=1241116664029303418, tenantId=1146029695717560320, journalId=1234093305789726721, articleId=1241116645033300311, language=EN, label=Fig.1, caption=Schematic diagram of PEMS installation, figureFileSmall=R+B+foXINdOvL5Qzy0+OHQ==, figureFileBig=eanG24aONmYrQXdyslb+mg==, tableContent=null), ArticleFig(id=1241116664167715462, tenantId=1146029695717560320, journalId=1234093305789726721, articleId=1241116645033300311, language=CN, label=图1, caption=PEMS安装示意, figureFileSmall=R+B+foXINdOvL5Qzy0+OHQ==, figureFileBig=eanG24aONmYrQXdyslb+mg==, tableContent=null), ArticleFig(id=1241116664410985134, tenantId=1146029695717560320, journalId=1234093305789726721, articleId=1241116645033300311, language=EN, label=Fig.2, caption=Real-world NH3emissions characteristic, figureFileSmall=iOgaKIh6YD+E3pmPpcLfuQ==, figureFileBig=qo7xz8+/i52ZLEpRtPDL7A==, tableContent=null), ArticleFig(id=1241116664515842744, tenantId=1146029695717560320, journalId=1234093305789726721, articleId=1241116645033300311, language=CN, label=图2, caption=实际道路行驶NH3排放特征, figureFileSmall=iOgaKIh6YD+E3pmPpcLfuQ==, figureFileBig=qo7xz8+/i52ZLEpRtPDL7A==, tableContent=null), ArticleFig(id=1241116664692003533, tenantId=1146029695717560320, journalId=1234093305789726721, articleId=1241116645033300311, language=EN, label=Fig.3, caption=Overall framework of the CNN-Transformer model, figureFileSmall=0fh/KjylWNvsEnmpByo+Kg==, figureFileBig=4QveB9mogIAHdfmv0IybZQ==, tableContent=null), ArticleFig(id=1241116664842998491, tenantId=1146029695717560320, journalId=1234093305789726721, articleId=1241116645033300311, language=CN, label=图3, caption=CNN-Transformer模型整体框架, figureFileSmall=0fh/KjylWNvsEnmpByo+Kg==, figureFileBig=4QveB9mogIAHdfmv0IybZQ==, tableContent=null), ArticleFig(id=1241116664998187757, tenantId=1146029695717560320, journalId=1234093305789726721, articleId=1241116645033300311, language=EN, label=Fig.4, caption=CNN block structure diagram, figureFileSmall=Dex0V/LUQFq5WAbP/JmGkw==, figureFileBig=lPQhVPuc3RbsrkCWm9zh1g==, tableContent=null), ArticleFig(id=1241116665119822595, tenantId=1146029695717560320, journalId=1234093305789726721, articleId=1241116645033300311, language=CN, label=图4, caption=CNN特征提取模块结构, figureFileSmall=Dex0V/LUQFq5WAbP/JmGkw==, figureFileBig=lPQhVPuc3RbsrkCWm9zh1g==, tableContent=null), ArticleFig(id=1241116665249846032, tenantId=1146029695717560320, journalId=1234093305789726721, articleId=1241116645033300311, language=EN, label=Fig.5, caption=Transformer block structure diagram, figureFileSmall=o6tHqPbETvoTe7UAZ0w/xg==, figureFileBig=99Sc/ftSByzz3oHZWeN4zg==, tableContent=null), ArticleFig(id=1241116665392452384, tenantId=1146029695717560320, journalId=1234093305789726721, articleId=1241116645033300311, language=CN, label=图5, caption=Transformer编码器结构图, figureFileSmall=o6tHqPbETvoTe7UAZ0w/xg==, figureFileBig=99Sc/ftSByzz3oHZWeN4zg==, tableContent=null), ArticleFig(id=1241116665497310001, tenantId=1146029695717560320, journalId=1234093305789726721, articleId=1241116645033300311, language=EN, label=Fig.6, caption=The main process of model building, figureFileSmall=9LND+9K8o7ApoeFpFSN0bg==, figureFileBig=e+yA9FfnVx/8/pyhYpSCoA==, tableContent=null), ArticleFig(id=1241116665623139135, tenantId=1146029695717560320, journalId=1234093305789726721, articleId=1241116645033300311, language=CN, label=图6, caption=模型构建过程, figureFileSmall=9LND+9K8o7ApoeFpFSN0bg==, figureFileBig=e+yA9FfnVx/8/pyhYpSCoA==, tableContent=null), ArticleFig(id=1241116665732191052, tenantId=1146029695717560320, journalId=1234093305789726721, articleId=1241116645033300311, language=EN, label=Fig.7, caption=Schematic of box plot method, figureFileSmall=/q/Z8yl5Oy2VwL0s5Oy/TQ==, figureFileBig=yiyayawTALx6qldayONFHw==, tableContent=null), ArticleFig(id=1241116667313443679, tenantId=1146029695717560320, journalId=1234093305789726721, articleId=1241116645033300311, language=CN, label=图7, caption=箱线图示意, figureFileSmall=/q/Z8yl5Oy2VwL0s5Oy/TQ==, figureFileBig=yiyayawTALx6qldayONFHw==, tableContent=null), ArticleFig(id=1241116667397329769, tenantId=1146029695717560320, journalId=1234093305789726721, articleId=1241116645033300311, language=EN, label=Fig.8, caption=Correlation between parameters, figureFileSmall=DdXct2EDK1GpC/cbgJYR9w==, figureFileBig=XkagGWgRNGaO4CWU+uAxpg==, tableContent=null), ArticleFig(id=1241116667502187383, tenantId=1146029695717560320, journalId=1234093305789726721, articleId=1241116645033300311, language=CN, label=图8, caption=各参数间相关性, figureFileSmall=DdXct2EDK1GpC/cbgJYR9w==, figureFileBig=XkagGWgRNGaO4CWU+uAxpg==, tableContent=null), ArticleFig(id=1241116667602850691, tenantId=1146029695717560320, journalId=1234093305789726721, articleId=1241116645033300311, language=EN, label=Fig. 9, caption=Schematic diagram of cosine annealing learning rate, figureFileSmall=pAUSqi8T1Ao0VZ9g5yEXAQ==, figureFileBig=b5GP5TR6xvK7L06RnC4SWw==, tableContent=null), ArticleFig(id=1241116667724485520, tenantId=1146029695717560320, journalId=1234093305789726721, articleId=1241116645033300311, language=CN, label=图9, caption=余弦退火学习率调整示意, figureFileSmall=pAUSqi8T1Ao0VZ9g5yEXAQ==, figureFileBig=b5GP5TR6xvK7L06RnC4SWw==, tableContent=null), ArticleFig(id=1241116667984532383, tenantId=1146029695717560320, journalId=1234093305789726721, articleId=1241116645033300311, language=EN, label=Fig.10, caption=Results of CNN-Transformer fusion model for predicting NH3 emissions, figureFileSmall=B0B+Man/aJ9E0EjfZ2QIyA==, figureFileBig=GOD9B+EIJ9IfoAJU9f3RYA==, tableContent=null), ArticleFig(id=1241116668139721648, tenantId=1146029695717560320, journalId=1234093305789726721, articleId=1241116645033300311, language=CN, label=图10, caption=CNN-Transformer融合模型预测结果分析, figureFileSmall=B0B+Man/aJ9E0EjfZ2QIyA==, figureFileBig=GOD9B+EIJ9IfoAJU9f3RYA==, tableContent=null), ArticleFig(id=1241116668244579257, tenantId=1146029695717560320, journalId=1234093305789726721, articleId=1241116645033300311, language=EN, label=Fig.11, caption=Comparison of NH3 emissions predictions for different models, figureFileSmall=4uLpfab12r9ZQ/GNA8soBw==, figureFileBig=JjZ5EcliQRcXRKLJ/P7T6A==, tableContent=null), ArticleFig(id=1241116668374602693, tenantId=1146029695717560320, journalId=1234093305789726721, articleId=1241116645033300311, language=CN, label=图11, caption=各模型NH3排放预测对比, figureFileSmall=4uLpfab12r9ZQ/GNA8soBw==, figureFileBig=JjZ5EcliQRcXRKLJ/P7T6A==, tableContent=null), ArticleFig(id=1241116668517209047, tenantId=1146029695717560320, journalId=1234093305789726721, articleId=1241116645033300311, language=EN, label=Fig.12, caption=Comparison of evaluation results for different models, figureFileSmall=RmEZiQ15fyllVHDaplrYpg==, figureFileBig=buqlj1ICIXZ0cVzLxkI4Lg==, tableContent=null), ArticleFig(id=1241116668647232483, tenantId=1146029695717560320, journalId=1234093305789726721, articleId=1241116645033300311, language=CN, label=图12, caption=各模型评价结果对比, figureFileSmall=RmEZiQ15fyllVHDaplrYpg==, figureFileBig=buqlj1ICIXZ0cVzLxkI4Lg==, tableContent=null), ArticleFig(id=1241116668819198963, tenantId=1146029695717560320, journalId=1234093305789726721, articleId=1241116645033300311, language=EN, label=Fig.13, caption=Feature importance analysis, figureFileSmall=OsP3p2IT3QaEe/DlHvd+hw==, figureFileBig=XhrEdrO/t/qP4hoZ2HIKvg==, tableContent=null), ArticleFig(id=1241116668932444160, tenantId=1146029695717560320, journalId=1234093305789726721, articleId=1241116645033300311, language=CN, label=图13, caption=特征重要性分析, figureFileSmall=OsP3p2IT3QaEe/DlHvd+hw==, figureFileBig=XhrEdrO/t/qP4hoZ2HIKvg==, tableContent=null), ArticleFig(id=1241116669066661900, tenantId=1146029695717560320, journalId=1234093305789726721, articleId=1241116645033300311, language=EN, label=Table 1, caption=

Optimization results of main hyperparameters

, figureFileSmall=null, figureFileBig=null, tableContent=
超参数设置
Conv1卷积核大小(Kernel Size)3
Conv2卷积核大小(Kernel Size)5
Transformer编码层数2
Transformer注意力头数6
FNN隐藏层维度102
初始学习率0.01
批量大小32
Dropout率0.3
L2正则化参数1×10-4
), ArticleFig(id=1241116669188296732, tenantId=1146029695717560320, journalId=1234093305789726721, articleId=1241116645033300311, language=CN, label=表1, caption=

模型关键超参数优化结果

, figureFileSmall=null, figureFileBig=null, tableContent=
超参数设置
Conv1卷积核大小(Kernel Size)3
Conv2卷积核大小(Kernel Size)5
Transformer编码层数2
Transformer注意力头数6
FNN隐藏层维度102
初始学习率0.01
批量大小32
Dropout率0.3
L2正则化参数1×10-4
)], attaches=null, journal=Journal(id=1234092555462295552, delFlag=0, nameCn=中国环境科学, nameEn=China Environmental Science, nameHistory1=null, nameHistory2=null, issn=1000-6923, eissn=null, cn=11-2201/X, coden=ZHKEEI, periodic=0, 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=fUkXgpzwRiw9vs+0dX4h8g==, journalPrice=null, startedYear=null, abbrevIsoEn=China Environmental Science, journalRemark=null, publicationField=null, createdTime=1772160193557, updatedTime=1772160729300, createdBy=18614031015, updatedBy=13701087609, firstLetterCn=C, firstLetterEn=C, subjectCode=Engineering, subjectName=null, subjectCodeEn=Engineering, subjectNameEn=null, picCn=fUkXgpzwRiw9vs+0dX4h8g==, picEn=w8+EIm00c59F/qhCr1EFJw==, jcr=null, cjcr=null, exts=[JournalExt(id=1234094802610999917, language=CN, name=中国环境科学, 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=1772160729315, updatedTime=1772160729315, createdBy=13701087609, updatedBy=13701087609, submissionGuidelinesUrl=, submissionAuthorUrl=https://zghjkxauthor.manuscriptcloud.com/, submissionEditorUrl=https://zghjkxeditor.manuscriptcloud.com/, submissionReviewUrl=https://zghjkxauthor.manuscriptcloud.com/, submissionCeEditorUrl=, submissionAeEditorUrl=, option={"copyright":""}), JournalExt(id=1234094802665525870, language=EN, name=China Environmental Science, 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=1772160729328, updatedTime=1772160729328, createdBy=13701087609, updatedBy=13701087609, submissionGuidelinesUrl=, submissionAuthorUrl=https://zghjkxauthor.manuscriptcloud.com/, submissionEditorUrl=https://zghjkxeditor.manuscriptcloud.com/, submissionReviewUrl=https://zghjkxauthor.manuscriptcloud.com/, submissionCeEditorUrl=, submissionAeEditorUrl=, option={"copyright":""})], databaseList=null, tenantJournalId=1234093305789726721, websiteList=[Website(id=1234095050196578613, webName=null, webTitle=null, webDomain=null, webCopyrigh=null, webIpcNo=null, seoTitle=null, seoKeywords=null, seoDescription=null, tenantJournalId=null, journalId=1234093305789726721, 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/zghjkx/CN, language=CN, createTime=1772160788344, createBy=18614031015, updateTime=1772160813480, updateBy=18614031015, name=中国环境科学-中文, tplId=1146099689490845704, title=中国环境科学, delFlag=0, indexPage=/home, props=[WebsiteProps(id=1234097146769756836, tenantId=1146029695717560320, journalId=null, journalGroupId=null, siteId=1234095050196578613, code=articleTextType, value=kx, createTime=1772161288206, updateTime=1772161288206, creator=18614031015, updator=18614031015), WebsiteProps(id=1234097146748785313, tenantId=1146029695717560320, journalId=null, journalGroupId=null, siteId=1234095050196578613, code=banner, value=null, createTime=1772161288201, updateTime=1772161288201, creator=18614031015, updator=18614031015), WebsiteProps(id=1234097146786534055, tenantId=1146029695717560320, journalId=null, journalGroupId=null, siteId=1234095050196578613, code=grayFlag, value=0, createTime=1772161288210, updateTime=1772161288210, creator=18614031015, updator=18614031015), WebsiteProps(id=1234097146740396704, tenantId=1146029695717560320, journalId=null, journalGroupId=null, siteId=1234095050196578613, code=logo, value=https://castjournals.cast.org.cn/joweb/zghjkx/CN/file/pic?fileId=MkE5LKk3Qw7XuYcjPisdew==, createTime=1772161288199, updateTime=1772161288199, creator=18614031015, updator=18614031015), WebsiteProps(id=1234097146799116969, tenantId=1146029695717560320, journalId=null, journalGroupId=null, siteId=1234095050196578613, code=minRunFlag, value=0, createTime=1772161288213, updateTime=1772161288213, creator=18614031015, updator=18614031015), WebsiteProps(id=1234097146761368227, tenantId=1146029695717560320, journalId=null, journalGroupId=null, siteId=1234095050196578613, code=picServerUrl, value=https://castjournals.cast.org.cn/joweb/zghjkx/CN/file/pic, createTime=1772161288204, updateTime=1772161288204, creator=18614031015, updator=18614031015), WebsiteProps(id=1234097146794922664, tenantId=1146029695717560320, journalId=null, journalGroupId=null, siteId=1234095050196578613, code=silenceFlag, value=0, createTime=1772161288212, updateTime=1772161288212, creator=18614031015, updator=18614031015), WebsiteProps(id=1234097146757173922, tenantId=1146029695717560320, journalId=null, journalGroupId=null, siteId=1234095050196578613, code=staticResourcePath, value=https://castjournals.cast.org.cn/joweb/cast_kjdb_cn_619/, createTime=1772161288203, updateTime=1772161288203, creator=18614031015, updator=18614031015), WebsiteProps(id=1234097146773951141, tenantId=1146029695717560320, journalId=null, journalGroupId=null, siteId=1234095050196578613, code=themeColor, value=null, createTime=1772161288207, updateTime=1772161288207, creator=18614031015, updator=18614031015), WebsiteProps(id=1234097146782339750, tenantId=1146029695717560320, journalId=null, journalGroupId=null, siteId=1234095050196578613, code=themeStyle, value=null, createTime=1772161288209, updateTime=1772161288209, creator=18614031015, updator=18614031015)]), Website(id=1234095050309824825, webName=null, webTitle=null, webDomain=null, webCopyrigh=null, webIpcNo=null, seoTitle=null, seoKeywords=null, seoDescription=null, tenantJournalId=null, journalId=1234093305789726721, 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/zghjkx/EN, language=EN, createTime=1772160788371, createBy=18614031015, updateTime=1772160830384, updateBy=18614031015, name=中国环境科学-英文, tplId=1146101810881728533, title=China Environmental Science, delFlag=0, indexPage=/home, props=[WebsiteProps(id=1234097176519955118, tenantId=1146029695717560320, journalId=null, journalGroupId=null, siteId=1234095050309824825, code=articleTextType, value=kx, createTime=1772161295299, updateTime=1772161295299, creator=18614031015, updator=18614031015), WebsiteProps(id=1234097176494789291, tenantId=1146029695717560320, journalId=null, journalGroupId=null, siteId=1234095050309824825, code=banner, value=null, createTime=1772161295293, updateTime=1772161295293, creator=18614031015, updator=18614031015), WebsiteProps(id=1234097176536732337, tenantId=1146029695717560320, journalId=null, journalGroupId=null, siteId=1234095050309824825, code=grayFlag, value=0, createTime=1772161295303, updateTime=1772161295303, creator=18614031015, updator=18614031015), WebsiteProps(id=1234097176486400682, tenantId=1146029695717560320, journalId=null, journalGroupId=null, siteId=1234095050309824825, code=logo, value=https://castjournals.cast.org.cn/joweb/zghjkx/EN/file/pic?fileId=MkE5LKk3Qw7XuYcjPisdew==, createTime=1772161295291, updateTime=1772161295291, creator=18614031015, updator=18614031015), WebsiteProps(id=1234097176545120947, tenantId=1146029695717560320, journalId=null, journalGroupId=null, siteId=1234095050309824825, code=minRunFlag, value=0, createTime=1772161295305, updateTime=1772161295305, creator=18614031015, updator=18614031015), WebsiteProps(id=1234097176511566509, tenantId=1146029695717560320, journalId=null, journalGroupId=null, siteId=1234095050309824825, code=picServerUrl, value=https://castjournals.cast.org.cn/joweb/zghjkx/EN/file/pic, createTime=1772161295297, updateTime=1772161295297, creator=18614031015, updator=18614031015), WebsiteProps(id=1234097176540926642, tenantId=1146029695717560320, journalId=null, journalGroupId=null, siteId=1234095050309824825, code=silenceFlag, value=0, createTime=1772161295304, updateTime=1772161295304, creator=18614031015, updator=18614031015), WebsiteProps(id=1234097176503177900, tenantId=1146029695717560320, journalId=null, journalGroupId=null, siteId=1234095050309824825, code=staticResourcePath, value=https://castjournals.cast.org.cn/joweb/cast_kjdb_en_623/, createTime=1772161295295, updateTime=1772161295295, creator=18614031015, updator=18614031015), WebsiteProps(id=1234097176524149423, tenantId=1146029695717560320, journalId=null, journalGroupId=null, siteId=1234095050309824825, code=themeColor, value=null, createTime=1772161295300, updateTime=1772161295300, creator=18614031015, updator=18614031015), WebsiteProps(id=1234097176528343728, tenantId=1146029695717560320, journalId=null, journalGroupId=null, siteId=1234095050309824825, code=themeStyle, value=null, createTime=1772161295301, updateTime=1772161295301, creator=18614031015, updator=18614031015)])], journalTitle=中国环境科学, weixinUrl=null, journalUrl=http://www.zghjkx.com.cn/, iacademicId=null, status=1, seqNo=null, journalTitleEn=China Environmental Science, journalPhotoCn=fUkXgpzwRiw9vs+0dX4h8g==, journalPhotoEn=w8+EIm00c59F/qhCr1EFJw==, journalFirstLetter=C, 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), detailUrlCn=https://castjournals.cast.org.cn/joweb/zghjkx/CN/Y2025/V45/I3/1231, detailUrlEn=https://castjournals.cast.org.cn/joweb/zghjkx/EN/Y2025/V45/I3/1231, pdfUrlCn=https://castjournals.cast.org.cn/joweb/zghjkx/CN/PDF/Y2025/V45/I3/1231, pdfUrlEn=https://castjournals.cast.org.cn/joweb/zghjkx/EN/PDF/Y2025/V45/I3/1231, aliStartDate=null, aliEndDate=null, collectionFlag=false, citedCount=null, citedUrl=null, reference=null)
收藏切换
基于CNN-Transformer融合框架的柴油车氨排放预测方法
收藏切换
PDF下载
白晓鑫 1, * , 郭向阳 1 , 吴春玲 1, 2 , 王凤滨 1 , 李旭 1 , 刘卫林 1
中国环境科学 | 大气污染与控制 2025,45(3): 1231-1240
收起
收藏切换
中国环境科学 | 大气污染与控制 2025, 45(3): 1231-1240
基于CNN-Transformer融合框架的柴油车氨排放预测方法
全屏
白晓鑫1, * , 郭向阳1, 吴春玲1, 2, 王凤滨1, 李旭1, 刘卫林1
作者信息
  • 1.中汽研汽车检验中心(天津)有限公司,天津 300300
  • 2.天津大学机械工程学院,天津 300072
  • 白晓鑫(1993-),男,山西阳泉人,工程师,硕士,主要从事重型发动机及整车排放控制研究.发表论文10余篇..

通讯作者:

* 责任作者,工程师,
Research on diesel vehicle NH3 emission prediction method based on CNN-Transformer fusion framework
Xiao-xin BAI1, * , Xiang-yang GUO1, Chun-ling WU1, 2, Feng-bin WANG1, Xu LI1, Wei-lin LIU1
Affiliations
  • 1.CATARC Automotive Test Center (Tianjin) Co., Ltd., Tianjin 300300, China
  • 2.School of Mechanical Engineering, Tianjin University, Tianjin 300072, China
出版时间: 2025-03-20
文章导航
收藏切换

本研究提出了一种基于卷积神经网络(CNN)和Transformer融合框架的柴油车NH3排放预测模型.该模型充分结合了CNN的局部特征提取能力和Transformer的全局依赖关注能力,实现了对实际行驶条件下柴油车NH3排放的高精度预测.研究以一辆N3类柴油车的实际道路排放测试数据为基础,采用Pearson相关系数法进行特征筛选,并利用贝叶斯优化算法对模型关键超参数进行调整,以提升模型性能.此外,应用SHAP算法量化了影响NH3排放的关键因素.结果表明,所提模型在独立数据集上能够高精度预测柴油车实际道路行驶中的NH3浓度排放,其预测的NH3浓度与实际测量值的R2、平均绝对误差(MAE)和均方误差(MSE)分别达到0.986、0.663和2.285,预测性能显著优于传统的随机森林(RF)模型、长短期记忆(LSTM)神经网络模型以及Transformer模型.研究为在用柴油车NH3排放监测提供了一种高效可靠的方法,同时为深入理解影响柴油车实际道路NH3排放的关键因素提供了新的研究思路.

柴油车  /  排放  /  NH3  /  卷积神经网络  /  Transformer

In this study, a diesel vehicle NH3 emission prediction model based on the fusion framework of Convolutional Neural Network(CNN)and Transformer is proposed. The model was developed by integrating the local feature extraction capability of CNN with the global dependency modeling capability of Transformer, enabling the highly accurate prediction of NH3 emissions from diesel vehicles under real road driving conditions. The study was conducted based on the actual on-road emissions test data of an N3-class diesel vehicle. Feature screening was performed using the Pearson correlation coefficient method, and the key hyperparameters of the model were optimized through the application of the Bayesian algorithm, which enhanced its performance. Additionally, the SHapley Additive exPlanations(SHAP)algorithm was utilized to identify the pivotal factors influencing NH3 emissions. The results indicated that the proposed model achieved highly accurate predictions of NH3 emissions from diesel vehicles in real road driving conditions when tested on an independent dataset. The R2, MAE, and MSE values of the predicted NH3 concentration compared to the actual measured values were 0.986, 0.663, and 2.285, respectively, which were significantly superior to those obtained by the traditional Random Forest(RF)model, the Long Short-Term Memory(LSTM)neural network model, and the Transformer model. This study provided an efficient and reliable method for monitoring NH3 emissions from in-use diesel vehicles and offered a novel perspective for elucidating the principal factors influencing NH3 emissions from diesel vehicles on the road.

diesel vehicles  /  emission  /  NH3  /  convolutional neural network  /  transformer
白晓鑫, 郭向阳, 吴春玲, 王凤滨, 李旭, 刘卫林. 基于CNN-Transformer融合框架的柴油车氨排放预测方法. 中国环境科学, 2025 , 45 (3) : 1231 -1240 .
Xiao-xin BAI, Xiang-yang GUO, Chun-ling WU, Feng-bin WANG, Xu LI, Wei-lin LIU. Research on diesel vehicle NH3 emission prediction method based on CNN-Transformer fusion framework[J]. China Environmental Science, 2025 , 45 (3) : 1231 -1240 .
城市化进程的加速与机动车保有量的急剧增长导致了一系列环境问题,其中柴油车排气污染物对空气质量的影响尤为显著,对公众健康构成了潜在威胁[1].在众多排气污染物中,NH3因其特殊的化学性质而备受关注.NH3不仅直接危害人体呼吸系统,更能通过复杂的大气化学反应参与细颗粒物(PM2.5)的生成过程,进而加剧雾霾天气的形成[2].近年来,多项研究表明[3-5],机动车已成为城市地区NH3排放的主要来源之一.值得注意的是,柴油车NH3排放主要来源于其尾气后处理系统的选择性催化还原(SCR)装置[6].SCR技术虽然有效降低了氮氧化物(NOx)的排放,但在工作过程中也会导致部分NH3的额外释放.因此,加强对柴油车NH3排放的监管变得愈发迫切.
为了有效控制柴油车NH3排放对环境的危害,全球主要法规体系已对此提出了严格的监管要求.我国现行的重型车国六排放标准GB17691-2018[7]明确规定了柴油发动机台架排放测试的NH3排放限值.然而,大量研究表明[8-10],实验室台架测量结果与真实道路条件下的NH3排放水平存在显著差异.相关研究[8]使用便携式傅里叶变换红外线光谱分析仪(FTIR)对一辆配备氧化性催化器(DOC)-SCR-颗粒捕集器(DPF)后处理系统的柴油车进行实际道路排放试验,结果显示,受环境温度、驾驶特征及SCR性能等因素影响,车辆在实际驾驶条件下的NH3排放因子约为实验室测试的2倍.鉴于此,2024年5月8日发布的欧七排放标准前瞻性地引入了便携式排放测试系统(PEMS),用于评估柴油车实际道路行驶NH3排放,并设定了具体限值要求[11].然而,PEMS测试是在环境条件、行驶工况比例等受控条件下进行的,难以全面反映车辆在多样化驾驶行为和工况下的NH3排放状况[12].此外,受试验成本和时间等因素限制,PEMS测试难以实现对在用柴油车NH3排放的全面监测.
鉴于现有监管措施的局限性,建立整车实际道路NH3排放预测模型成为NH3排放监管的重要研究方向.然而,准确预测柴油车NH3排放仍面临诸多挑战.首先,NH3排放受发动机工况、SCR系统性能以及环境温度等多因素影响,这些因素间存在复杂的非线性关系[13].其次,传统的排放预测方法往往依赖于简化模型或经验公式,难以全面捕捉NH3排放的动态特性[14-15].近年来,深度学习方法在排放预测领域展现出巨大潜力.CNN因其强大的特征提取能力,已被广泛应用于时序数据分析[16].而Transformer模型凭借其优异的长程依赖建模能力,在序列预测任务中取得了显著成果[17].然而,单一模型往往难以同时兼顾局部特征提取和全局依赖建模,限制了其在复杂排放预测任务中的表现.为解决上述问题,本研究提出了一种创新的CNN-Transformer融合网络框架,用于柴油车实际道路NH3排放预测.通过实车测试数据,验证了所提方法的有效性和优越性,以期为未来柴油车NH3排放控制与监管提供科学依据和方法参考.
为探究柴油车在实际道路行驶条件下的NH3排放特征,为构建准确、可靠的排放预测模型奠定坚实的数据基础,本研究选取一辆符合国六排放标准的N3类重型柴油车进行了实际道路排放测试.测试车辆配备了一台排量为5.193L、额定功率为165kW的柴油发动机,采用了DOC+DPF+SCR的后处理技术组合.
测试采用HORIBA公司研发的OBS-ONE车载排放测试系统.该系统主要由排气污染物测量模块、排气流量计、数据通讯模块和环境监控模块组成,能够实时获取车辆运行状态、尾气中主要污染物浓度、排气流量、外界环境参数和车辆位置信息等数据.同时,该系统与车载诊断(OBD)接口连接,可实时读取发动机及尾气后处理装置的关键运行参数,如发动机指示转矩、摩擦转矩和转速等.使用OBS-ONE-XL量子级联激光器红外光谱仪(QCL-IR)测量尾气中的NH3浓度,其有效测量范围覆盖0~1500×10-6,体积分数.经验证,QCL-IR的测量结果与实验室固定式傅里叶变换红外光谱仪设备测量结果[8]具有非常好的一致性.这些多维数据为全面分析NH3排放特征及构建预测模型提供了可能.
图2呈现了冷启动条件下测试车辆在实际道路行驶期间的瞬时NH3排放特征.结果显示,在整个市区工况期间,由于车辆SCR系统未达到有效工作温度,NH3排放基本为零.随着车辆进入市郊路段,发动机负荷增加导致排气温度上升.当排气温度达到尿素水解温度时,SCR系统开始正常工作,尾气中随即检测到NH3排放.值得注意的是,市郊工况下的NH3排放水平显著低于高速工况.这主要是因为市郊工况下发动机负荷较低,导致排气温度和NOx排放量较低,从而减少了SCR系统中NH3的生成和泄漏.
图2所示为行驶条件对NH3排放的显著影响.在动态驾驶和加速等瞬态工况下,NH3排放量明显增加.这可能是由于瞬态工况下NOx排放急剧增加,以及SCR控制系统难以精确调节尿素喷射量所致.这一观察结果与先前的研究发现一致[10],进一步证实了实际道路行驶条件对柴油车NH3排放的重要影响.
基于上述分析,本研究在选择NH3排放预测模型的输入特征时,重点考虑了以下几个方面:(1)排气温度:直接影响SCR系统的NOx转化效率及NH3的泄漏.(2)发动机负荷和NOx排放量:与NH3的产生和泄漏密切相关.(3)车辆行驶工况相关参数:反映实际道路行驶条件对NH3排放的影响.
本文提出了一种新型柴油车NH3排放预测模型,模型利用了CNN和Transformer的互补优势,其结构如图3所示.该模型主要由3个关键组件构成:CNN特征提取模块、Transformer序列处理模块和输出层.首先,输入的时间序列数据通过CNN模块进行处理,以提取空间特征向量.为了进一步提升模型的特征提取能力,在设计中堆叠了多个卷积层.这种深层结构使得模型能够更全面地学习和理解复杂的实际道路NH3排放特性.随后,这些提取的特征被输入到具有多头自注意力机制的Transformer模块中,以充分挖掘时间维度的信息,同时获取潜在的时空相关性.最后,通过全连接层将Transformer的输出映射到一维输出,即模型预测的NH3浓度.
采用CNN专注对输入数据进行局部特征提取.CNN架构通常包含卷积、池化和全连接等关键层级[18].其中,卷积层作为CNN的核心组件,通过特定的数学运算从输入数据中捕获空间结构信息.具体而言,卷积核在输入数据上移动,对感受野内的数据执行点积运算并加入偏差,从而生成特征图.这一过程可用式(1)表示.池化层位于卷积层之后,用于对特征图进行降维,有效减少网络参数量和计算复杂度.网络末端的全连接层则将前述提取的特征进行展平和非线性转换,最终输出分类结果.
式中:yi为输出特征图中位置i的值;f为激活函数,用于卷积运算后进行非线性变换;D为卷积核的大小;Xi+j为输入序列中从位置i开始的一个局部窗口;j为卷积核内的位置索引;Wj为卷积核中位置j的值;b为偏置.
本研究构建的CNN模块结构如图4所示.该模块采用了双层一维卷积架构,通过多层卷积和池化操作,能够有效捕获数据中的短程空间依赖关系和局部特征,为后续的时序建模提供丰富的特征表示.模块的工作流程如下:首先,通过第一卷积层对输入数据进行初步特征提取.使用修正线性单元(ReLU)激活函数引入非线性变换,增强模型的表达能力.随后,第二卷积层对初级特征进行进一步提取,捕获更复杂的数据模式.再次应用ReLU函数,强化特征的非线性表示.这种双层卷积结构显著扩大了模块的感受野,有助于从输入数据中提取更丰富、更具代表性的特征.为了优化特征表示并增强模型的泛化能力,本模块还包含了最大池化层(Max-Pooling)和批量归一化(BN)层等关键组件.最大池化层用于降低特征维度,同时保留最显著的特征信息,提高计算效率和模型的空间不变性.而BN层用于特征标准化处理,这有助于加快训练速度并缓解梯度消失等深度学习中常见的问题[19].
虽然CNN在提取数据局部特征方面表现出色,但在处理具有长时间依赖关系的柴油车实际道路运行数据时,其效果往往不尽如人意.相比之下,Transformer作为当代机器学习领域最强大的模型之一,凭借其独特的结构设计,能够有效捕捉数据中的长程依赖关系,从而获得更深入的特征表示[20].因此,本研究引入Transformer编码器模块,使得模型可以兼顾全局特征和各数据维度的局部特征,显著增强模型对长距离依赖关系的感知能力,从而有效提升NH3排放预测的准确性.
本研究采用的Transformer编码器层主要由两个核心组件构成:多头自注意力机制(MSA)和前馈神经网络(FNN).此外,为了优化模型训练过程和提高泛化能力,引入了层归一化(LN)和Dropout机制.整体结构如图5所示.
自注意力机制是Transformer网络的核心创新,其设计灵感来源于人类选择性关注信息的认知过程.这一机制使得模型能够自适应地将计算资源分配到输入序列中最关键的部分,从而提高信息提取的效率和质量,进而改善模型的整体性能.自注意力机制的计算过程可以概括为以下步骤:(1)输入转换:将输入序列X通过三个不同的线性变换,分别生成查询(Q)、键(K)和值(V)向量.这一过程可表示为式(2);(2)注意力分数计算:通过QK的点积运算,计算注意力分数,并进行缩放以稳定梯度;(3)权重分配:使用softmax函数将注意力分数转换为概率分布,作为各个值向量的权重;(4)加权求和:将权重与V进行点乘,得到最终的注意力输出.这一过程可以用式(3)概括.
式中:WQWKWV分别是对应的权重矩阵;dk是键向量的维度.
Transformer进一步将自注意力机制扩展为多头注意力机制.通过将QKV向量分别映射到多个子空间,并在每个子空间独立计算注意力,最后将结果拼接并线性变换,使得模型能够同时关注输入序列的不同特征和位置.多头注意力的计算可表示为:
式中:headi表示第i个头的自注意力分布;表示第i个头的线性投影参数矩阵;WO表示输出投影的参数矩阵.
最后,通过一个线性层将Transformer编码器的高维特征表示映射到NH3排放预测值.这一输出层采用线性变换实现,其数学表达式如下:
式中:W表示权重矩阵;b为偏置向量;x为Transformer编码器的输出向量;y为最终预测的NH3排放值.
所提融合框架将CNN的局部特征提取和Transformer的全局依赖建模相互补充,实现了从局部到全局的多尺度特征提取,能够更全面地描述和预测复杂的NH3排放.
本研究的模型构建基于对测试车辆进行的多次实际道路排放试验数据.为有效捕捉车辆运行工况及排放特征,同时平衡数据采集量和设备性能,研究采用1Hz的采样率,共收集了23705组样本.这些数据全面涵盖了重型车辆在各种典型环境及工况下的性能表现,具体包括:(1)动态工况:涵盖加速、减速、匀速和怠速等不同行驶状态;(2)道路环境:包括市区路、市郊路和高速路等多种路况;(3)环境温度:覆盖低温(<0℃)、常温(0~25℃)和高温(>25℃)等不同气候条件.
本研究采用分层随机抽样方法,按照6:2:2的比例将数据集划分为训练数据集、验证数据集和测试数据集,以确保模型训练的有效性和评估结果的可靠性.
图6展示了模型构建的主要流程.首先,对原始数据进行预处理,包括缺失值和异常值处理.随后,通过特征选择确定对NH3排放预测最为关键的输入变量.为确保模型训练的稳定性和收敛性,对选定的输入特征进行标准化处理.经过上述预处理步骤的数据随即输入到本研究提出的CNN-Transformer融合框架中进行模型训练和参数优化.为了增强模型的可解释性,采用SHAP算法对模型的预测结果进行了深入分析,以揭示不同特征对NH3排放的贡献程度.最后,为了验证所提出方法的有效性和性能优势,选择了多个在时间序列预测领域广泛应用的基准模型进行了对比分析.
数据预处理是确保模型性能的关键步骤,本研究主要进行了异常值检测与处理.采用箱线图法进行异常值检测[21],将小于Q1-1.5*IQR或大于Q3+1.5*IQR的数据定义为潜在异常值,如图7所示.其中,Q1、Q3分别为数据的第25%和第75%分位值,IQR为四分位距(Q3-Q1).对于检测到的潜在异常值,结合专业知识进行进一步的验证.考虑到试验数据量有限,将确认的异常值视为缺失值处理,以避免直接删除可能导致的信息损失.采用K近邻(KNN)插补法填补缺失值,以最大限度地保留数据信息.KNN方法通过在缺失值附近寻找若干最相似的历史数据来估算并填补缺失值,能够在保持数据分布特征的同时,有效处理多变量数据集中的缺失值问题[22].具体的KNN插补步骤如下:(1)对于每个包含缺失值的样本,根据欧氏距离等度量方法选择距离缺失样本最近的K个样本(本研究中K=5);(2)使用这K个样本数据的加权平均值来填补缺失值,权重与样本间距离成反比.
基于前文实际道路NH3排放特征分析和现有研究[23,8]对NH3排放影响因素的深入探讨,本研究初步选择了13个潜在特征用于回归建模.这些特征涵盖了车辆运行状态、发动机参数和后处理系统性能等多个方面,包括:车速、发动机转速、发动机指示扭矩、摩擦扭矩、进气量、燃料流量、排气流量、冷却液温度、SCR入口排气温度、SCR出口排气温度、SCR上游NOx浓度、SCR下游NOx浓度和NH3浓度.
为分析特征间的相互关系,以进一步优化模型输入,对选定的特征进行了皮尔逊(Pearson)相关性分析,计算公式如下:
式中:X1X2分别表示两个特征数据;μX1μX2为特征X1X2的平均值;σX1σX2为特征X1X2的方差;E为数学期望计算.
图8展示了不同特征间的相关性矩阵.由图可以看出,进气量、燃料流量和排气流量之间存在高度相关性.同时,SCR入口排气温度和SCR出口排气温度也表现出较强的线性关系.因此,在特征工程中删除了SCR出口排气温度、进气量和燃料流量,以减少特征可能提供的冗余信息,降低模型复杂度.
为降低输入特征间量纲差异对模型性能的影响,并加快网络权重参数的收敛,本研究采用Z-score标准化方法对训练集输入特征数据进行预处理.这种方法将每个特征转换为均值为0、标准差为1的标准正态分布数据,既保持了原始数据的分布特征,又使不同量纲的特征具有可比性.标准化处理的数学表达式如下:
式中:xnorm,i为标准化后的第i个数据;xi为原始的第i个数据;为特征数据的平均值;σ为特征数据的标准差.为防止数据泄露并确保模型泛化能力的准确评估,使用训练集得到的标准化参数对测试集进行相同的标准化处理.
考虑到NH3排放具有明显的时序依赖特性,研究采用滑动窗口技术生成时序样本,以充分利用数据的时序信息,提高模型的预测精度.其中,滑动窗口的大小选择为10,步长为1.样本生成过程可以表示为:
式中:Xi是第i个的输入特征向量(包含前9个时间步和当前时间步的所有特征);yi为第i个NH3浓度值.
模型训练采用交叉熵损失作为损失函数,并利用AdamW优化器更新模型的可学习参数[24].AdamW优化器结合了Adam优化器和权重衰减,不仅具有自适应学习率的优势,还能通过权重衰减有效防止过拟合.采用带热重启的余弦退火策略(Cosine annealing)对学习率进行调整[25],以确保训练过程中的平稳收敛和性能提升.余弦退火策略按照式(10)调整学习率,学习率调整结果如图9所示.
式中:ηt为当前步骤的学习率;ηmin为设定的最小学习率;ηmax为设定的最大学习率;t为当前训练步骤;T为总训练步骤.
此外,在训练过程中引入了早停策略,并采用了L2正则化和丢弃率(Dropout)技术,以控制模型复杂度并增强模型的泛化能力.
在模型构建过程中,优化算法超参数是确保模型性能优越的关键步骤.本研究采用了贝叶斯优化算法[26]结合K折交叉验证[27]的方法来调整模型超参数,以获得最佳预测性能.表1详细列出了CNN-Transformer模型关键超参数的优化结果.
为全面评估模型的预测性能,本研究采用3个互补的评估指标:决定系数(R2)、平均绝对误差(MAE)和均方误差(MSE).其中,R2反映模型预测结果与实际测量值的整体吻合程度,R2越接近1,则自变量对因变量的解释能力越强.MAE用于衡量模型预测的平均偏离程度,其对异常值不敏感,可以提供稳健的误差估计.MSE通过计算预测误差的平方平均值量化预测值与真实值间的偏差.相比MAE,MSE更能识别显著偏差.MAE与MSE越小,则表明模型的预测准确性越高.综合使用这3个指标可以从不同角度评估模型性能,既考虑整体拟合程度,又关注误差的大小和分布特征.这些评价指标的具体计算方法分别见式(11)~(13).
式中:为NH3排放第i个预测值,10-6yi为NH3排放第i个实际测量值,10-6;为全部NH3排放测量值的平均值,10-6n为NH3排放测量值的数量.
为深入理解各特征对NH3排放预测的影响,本研究利用SHAP算法对模型进行了可解释性分析,为揭示模型决策的内在机制和制定有效的排放控制策略提供了有益参考.对于给定的预测实例,SHAP算法衡量了在所有可能的特征组合中,引入某个特征对模型输出的平均影响[28].SHAP值的大小反映了特征对模型决策的重要程度.这种方法不仅考虑了特征的独立作用,还捕捉了特征间的交互效应,从而提供了全面而准确的解释.
为验证所提出的CNN-Transformer融合框架在柴油车实际道路NH3排放预测中的优势,本研究选择了时序数据预测领域常用的LSTM和RF模型进行对比实验.同时,为评估CNN模块在所提模型中的作用,还进行了使用单独Transformer模型的消融实验.在确保所有模型超参数均已优化的基础上进行了对比实验和消融实验.此外,所有模型使用训练数据集和测试数据集相同,并通过前文提及的3个评价指标对各模型的性能进行定量评估.
图10展示了优化后的CNN- Transformer融合框架在独立测试数据集上的预测性能.如图10(a)所示,绝大多数数据点紧密围绕着理想拟合线y=x分布,这表明模型预测值与QCL-IR仪器的测量值之间存在极高的线性相关性,其决定系数R2高达0.986.图10(b)展示了部分模型预测值与QCL-IR仪器测量值的时间序列对比,两条曲线的高度重合进一步验证了CNN-Transformer融合模型在重型车辆NH3排放预测方面的优异性能.
为了深入评估各模型的预测效果,选取了具有代表性的时段进行重点分析,结果如图11所示.分析表明,RF模型在预测测试车辆实际道路行驶NH3排放时表现欠佳,预测值与实际值存在明显偏差.LSTM模型相较于RF模型有所改善,预测曲线整体趋近于真实值,但在低浓度排放数据上仍存在一定误差.
特别值得注意的是,NH3浓度预测误差主要出现在大多数的波峰和波谷等特殊工况区域.在这些区域,所提出的CNN-Transformer融合相比于消融模型Transformer表现最为出色,能更准确地捕捉NH3排放的动态特征.这主要得益于CNN对输入数据局部特征的有效提取,有效降低了由于柴油车辆运行条件的急剧变化所引起的预测偏差.
为了量化评估各模型在测试集上的预测效果,采用前文选择的3个评估指标进行比较,结果如图12所示.由图可以看出,所提出的CNN-Transformer融合模型在测试集上的表现优异,其NH3浓度预测值与真实值的R2、MAE和MSE分别为0.986、0.663和2.285.与RF、LSTM和Transformer模型相比,所提模型预测结果的R2分别提高了8.30%、4.49%和2.55%;MAE分别降低了70.50%、60.99%和56.47%;MSE分别降低了84.64%、75.54%和64.15%.这充分表明,相比于传统的Transformer、LSTM和RF模型,所提出的CNN-Transformer融合模型能更有效地预测柴油车实际道路行驶中的NH3排放,预测精度显著提升.
图13为各特征对模型预测结果的贡献度.结果表明,与SCR系统相关的特征对NH3排放预测起着至关重要的作用,特别是SCR入口排气温度和SCR下游NOx浓度.这一发现与SCR系统的工作原理高度契合.SCR系统通过尿素水溶液热解产生的NH3来还原尾气中的NOx.SCR反应效率对温度条件极为敏感:过低的温度会导致尿素不完全分解,引发NH3泄漏;而过高的温度则可能促使NH3发生二次氧化,生成NOx.因此,SCR入口温度成为控制NH3排放的决定性因素.同时,SCR下游NOx浓度作为SCR系统还原效率的直接指标,反映了尿素热解产生的NH3利用程度,为NH3排放预测提供了至关重要的信息.这两个参数的高贡献度表明了精确控制SCR系统运行条件对减少柴油车实际道路NH3排放的重要性.
值得注意的是,车速和发动机输出扭矩等动态运行参数的SHAP值也相对较高,表明这些因素对准确预测NH3排放同样具有重要影响.这主要是因为这些参数直接决定了发动机负荷和排气动态特性,进而影响SCR系统的工作状态和NH3排放模式.这一发现强调了在实际道路工况下,考虑车辆动态运行特性对准确预测和控制NH3排放的必要性.
SHAP分析结果不仅验证了模型捕捉到了NH3排放机理的关键因素,还为优化SCR系统设计和控制策略提供了有力支持.例如,可以基于SCR入口温度和下游NOx浓度开发更精细的尿素喷射控制策略,或者结合车速和发动机扭矩信息,设计适应不同工况的智能排放控制系统.
未来研究中将进一步扩大数据样本范围,涵盖不同车型、后处理技术和使用年限的车辆,跨车辆的模型性能评估,以增强模型的普适性和泛化能力.同时,关注评估模型在长期使用过程中的性能变化.通过进行长期跟踪试验,评估模型在车辆老化、SCR系统性能退化等情况下的预测能力,并探索模型的自适应更新机制.
5.1 所提出的CNN-Transformer融合框架通过结合多层CNN和Transformer编码器,并结合贝叶斯算法和K折交叉验证进行超参数精细调优,实现了对柴油车实际道路行驶NH3排放的高精度预测.在测试集上,该模型预测的NH3浓度与实际测量值表现出了良好的一致性,R2达到0.986,MAE和MSE分别为0.663和2.285.
5.2 为评估模型性能,在一致的实验条件下构建了多种预测模型,并进行了对比实验与消融试验.结果表明,CNN-Transformer融合模型在关键评价指标上显著超越了其他模型.与次优的Transformer模型相比,所提模型R2提升2.55%,而MSE和MAE分别降低64.15%和56.47%,这表明所提模型在捕捉NH3排放复杂动态特征方面的显著优势,尤其是在处理时序数据和长期依赖关系方面的能力.
5.3 通过SHAP方法的可解释性分析,揭示了影响柴油车实际道路行驶NH3排放的关键因素.结果表明,SCR入口排气温度和SCR下游NOx浓度对NH3排放预测贡献最大,这与SCR系统的工作原理一致.同时,车速和发动机输出扭矩等动态运行参数也显示出较高的重要性,表明考虑实际道路工况对准确预测NH3排放的必要性.这些发现为优化SCR系统控制策略和减少NH3排放提供了重要依据.
  • 国家重点研发计划(2022YFC3701800)
参考文献 引证文献
排序方式:
[1]
中国移动源环境管理年报(2023年) [J]. 环境保护202452(2):48-62.
China mobile source environmental management annual report (2023)[J]. Environmental Protection202452(2):48-62.
[2]
Anderson NStrader RDavidson C. Airborne reduced nitrogen: Ammonia emissions from agriculture and other sources [J]. Environment international200329(2/3): 277-286.
[3]
Zhang HYing Q. Source apportionment of airborne particulate matter in Southeast Texas using a source-oriented 3D air quality model [J]. Atmospheric Environment201044(29):3547-3557.
[4]
Farren, N J.Davison JRose R A,et al. Underestimated ammonia emissions from road vehicles [J]. Environmental science & technology202054(24):15689-15697.
[5]
陈培林,肖欣欣,王勤耕. 基于卫星观测的2010~2020年中国高分辨率NH3排放特征 [J]. 中国环境科学202343(6):2673-2682.
Chen P LXiao X XWang Q G,et al. High-resolution characteristics of NH3 emission from 2010 to 2020 in China based on satellite observation [J]. China Environmental Science202343(6):2673-2682.
[6]
白晓鑫,吴春玲,刘卫林,等. 柴油车尿素溶液品质在线检测方法研究. 汽车实用技术202449(12):89-94.
Bai X XWu C LLiu W L,et al. Research on online detection method for urea solution quality of diesel vehicle [J]. Automobile Applied Technology202449(12):89-94.
[7]
GB 17691—2018 重型柴油车污染物排放限值及测量方法(中国第六阶段) [S].
GB 17697-2018 Limits and measurement methods for emissions from diesel-fueled heavy-duty vehicles (China VI) [S].
[8]
Suarez-Bertoa RMendoza-Villafuerte PRiccobono F,et al. On-road measurement of NH3 emissions from gasoline and diesel passenger cars during real world driving conditions [J]. Atmospheric Environment2017166:488-497.
[9]
Zhu HMa TToumasatos Z,et al. On-road NOx and NH3emissions measurements from in-use heavy-duty diesel and natural gas trucks in the South Coast air Basin of California [J]. Atmospheric Environment2024316:120179.
[10]
Mendoza-Villafuerte PSuarez-Bertoa RGiechaskiel B,et al. NOx,NH3,N2O and PN real driving emissions from a Euro VI heavy-duty vehicle. Impact of regulatory on-road test conditions on emissions [J]. Science of the Total Environment2017609:546-555.
[11]
European Commission. Proposal for a regulation of the European parliament and of the council on type-approval of motor vehicles and of engines and of systems,components and separate technical units intended for such vehicles,with respect to their emissions and battery durability (Euro 7) [EB/OL]. (2022-11-10)[2024-06-04]. https://eur-lex.europa.eu/legal-content/EN/TXT/?uri=CELEX%3A52022PC0586.
[12]
Kousoulidou MGeorgios FLeonidas N,et al. Use of portable emissions measurement system (PEMS) for the development and validation of passenger car emission factors [J]. Atmospheric Environment201364:329-338.
[13]
Thiruvengadam ABesch M CThiruvengadam P,et al. Emission rates of regulated pollutants from current technology heavy-duty diesel and natural gas goods movement vehicles [J]. Environmental Science & Technology201549(8):5236-5244.
[14]
Huang CHu QLou S,et al. Ammonia emission measurements for light-duty gasoline vehicles in China and implications for emission modeling [J]. Environmental Science & Technology201852(19):11223-11231.
[15]
Pla BPiqueras PBares P,et al. NOx sensor cross sensitivity model and simultaneous prediction of NOx and NH3 slip from automotive catalytic converters under real driving conditions [J]. International Journal of Engine Research202122(10):3209-3218.
[16]
Wen LLi XGao L,et al. A new convolutional neural network-based data-driven fault diagnosis method [J]. IEEE Transactions on Industrial Electronics201765(7):5990-5998.
[17]
Vaswani AShazeer NParmar N,et al. Attention is all you need [J]. Advances in Neural Information Processing Systems2017:5998-6008.
[18]
黄茂庭,徐金明. 使用CNN(卷积神经网络)-LSTM(长短期记忆)联合神经网络预测盾构隧道施工引起的地面沉降 [J]. 城市轨道交通研究202427(6):166-171.
Huang M TXu J M. Prediction of land subsidence caused by shield tunnel construction with joint CNN-LSTM neural network [J]. Urban Mass Transit202427(6):166-171.
[19]
Thakkar VTewary SChakraborty C. Batch Normalization in convolutional neural networks—A comparative study with CIFAR-10 data [C]//2018fifth international conference on emerging applications of information technology (EAIT). IEEE,2018:1-5.
[20]
Vaswani AShazeer NParmar N,et al. Attention is all you need [J]. Advances in Neural Information Processing Systems201730:5998-6008.
[21]
Williamson D FParker R AKendrick J S. The box plot: a simple visual method to interpret data [J]. Annals of internal medicine1989110(11):916-921.
[22]
陈婉娇. 缺失数据插补方法及其在医学领域的应用研究 [D]. 广州:华南理工大学,2020.
Chen W J. Research on application of missing data lmputation in medical field [D]. Guang Zhou: South China University of Technology,2020.
[23]
He LLi GWu X,et al. Characteristics of NOx and NH3 emissions from in-use heavy-duty diesel vehicles with various aftertreatment technologies in China [J]. Journal of Hazardous Materials2024465:133073.
[24]
Chen XLiang CHuang D,et al. Evolved optimizer for vision[C]//First Conference on Automated Machine Learning (Late-Breaking Workshop). 2022.
[25]
Loshchilov IHutter F. SGDR: Stochastic gradient descent with warm restarts [J]. 2016.
[26]
白晓鑫,吴春玲,景晓军,等. 基于RLS和BO算法的重型车载重估算研究 [J]. 汽车实用技术202348(5):56-63.
Bai X XWu C LJing X J,et al. Research on heavy-duty vehicle mass estimation based on recursive least square and bayesian optimization algorithm [J]. Automobile Applied Technology202348(5):56-63.
[27]
Jung Y. Multiple predicting K-fold cross-validation for model selection [J]. Journal of nonparametric statistics201830(1):197-215.
[28]
Marcílio W EEler D M. From explanations to feature selection: assessing SHAP values as feature selection mechanism[C]//2020 33rd SIBGRAPI conference on Graphics,Patterns and Images (SIBGRAPI).IEEE,2020: 340-347.
2025年第45卷第3期
PDF下载
44
19
引用本文
BibTeX
文章信息
  • 接收时间:2024-08-20
  • 首发时间:2026-03-18
  • 出版时间:2025-03-20
补充材料
相关文章
文章信息
作者
出版历史
  • 收稿日期:2024-08-20
基金
国家重点研发计划(2022YFC3701800)
作者信息
    1.中汽研汽车检验中心(天津)有限公司,天津 300300
    2.天津大学机械工程学院,天津 300072

通讯作者:

* 责任作者,工程师,
参考文献
分享链接
https://castjournals.cast.org.cn/joweb/zghjkx/CN/1241116645033300311
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