电源学报

基于双尺度卷积神经网络的TMMC直流母线电容参数辨识

PDF下载

代子阔 ¹ , 张东升 ² , 刘尚坤 ³ , 郭任 ⁴ , 陈丽娟 ⁵ , 汪诚 ⁶

电源学报 | 人工智能与电力电子 2026,24(4): 271-283

收起

电源学报 |人工智能与电力电子 2026 , 24 (4) : 271 -283

基于双尺度卷积神经网络的TMMC直流母线电容参数辨识

全屏

代子阔¹, 张东升², 刘尚坤³, 郭任⁴, 陈丽娟⁵, 汪诚⁶

作者信息

^1.国网辽宁省电力有限公司，沈阳　110000

^2.国网辽宁省电力有限公司鞍山供电公司，鞍山　114000

^3.国网阜新供电公司，阜新　123100

^4.国网沈阳供电公司，沈阳　110003

^5.国网本溪供电公司，本溪　117020

^6.南京理工大学自动化学院，南京　210094

通讯作者:

汪诚（1988—　），男，通信作者，博士，副教授。研究方向：模块化电力电子变换器，主电路拓扑与性能控制。E-mail：chw714@njust.edu.cn。

作者简介:

代子阔（1985—　），男，硕士，高级工程师。研究方向：智能配电网关键技术。E-mail：dzk19851022@ln.sgcc.com.cn。

张东升（1988—　），男，本科，工程师。研究方向：配网运维及供电可靠性管理。E-mail：zhangdongsheng_1@ln.sgcc.com.cn。

刘尚坤（1988—　），男，本科，工程师。研究方向：配电网运维管理。E-mail：liushangkun@ln.sgcc.com.cn。

郭任（1989—　），男，本科，助理工程师。研究方向：配电线路运维。E-mail：guoren_1@ln.sgcc.com.cn。

陈丽娟（1981—　），女，本科，高级工程师。研究方向：可靠性管理技术。E-mail：clj1981@ln.sgcc.com.cn。

Parameter Identification for DC-link Capacitors in TMMC Based on Double-scale Convolutional Neural Network

Zikuo DAI¹, Dongsheng ZHANG², Shangkun LIU³, Ren GUO⁴, Lijuan CHEN⁵, Cheng WANG⁶

Affiliations

^1.State Grid Liaoning Electric Power Co., Ltd., Shenyang 110000, China

^2.Anshan Power Supply Company, State Grid Liaoning Electric Power Co., Ltd., Anshan 114000, China

^3.State Grid Fuxin Power Supply Company, Fuxin 123100, China

^4.State Grid Shenyang Power Supply Company, Shenyang 110003, China

^5.State Grid Benxi Power Supply Company, Benxi 117020, China

^6.School of Automation, Nanjing University of Science and Technology, Nanjing 210094, China

出版时间: 2026-04-30 doi: 10.13234/j.issn.2095-2805.2026.4.271

文章导航

摘要

收起

T型级联模块化多电平变换器TMMC（T-cascade modular multilevel converter）结合多电平和级联模块化技术，具有九电平端口电压、冗余开关模式、多电源端口和硬件复用等特点，是实现高功率密度高效率不间断电源UPS（uninterruptible power supply）的理想拓扑之一。然而，直流母线电容器数量多、易失效，对TMMC-UPS的可靠运行构成威胁，因此，考虑对TMMC-UPS场合母线电容器进行监测来智能感知电容容量变化、预防电容退化故障，保证不间断供电可靠性。提出1种双尺度卷积神经网络模型，通过提取电容电压纹波的低频和中频特征，实现电容容值和串联电阻的双重识别，提高电容器状态监测精度。所提方法基于数据驱动，不依赖数学模型，同时不需要增加额外测量硬件，易推广到其他类型变换器的参数辨识领域。

关键词

电容 / T型级联模块化多电平变换器 / 参数辨识 / 双尺度卷积神经网络

Abstract

收起

By combing the multi-level technology with the cascade modular technology, the T-cascade modular multilevel converter (TMMC) has characteristics such as 9-level port voltage, redundant switching mode, multiple power ports and hardware multiplexing, and it has become one of the ideal topologies for realizing uninterruptible power supply (UPS) with high power density and high efficiency. However, the large number of DC-link capacitors which are susceptible to failures poses a threat to the reliable operation of TMMC-UPS. Therefore, it is necessary to monitor DC-link capacitors under the scenario of TMMC-UPS to intelligently perceive the changes in the capacitance value and prevent the degradation failure of capacitors, so as to ensure the reliability of UPS. Therefore, a double-scale convolutional neural network model is proposed. By extracting the low- and medium-frequency features of capacitor voltage ripple, both the capacitance value and the equivalent series resistance can be identified, thus improving the monitoring accuracy of capacitor status. The proposed data-driven method does not rely on any mathematical model, and it does not need to add additional measurement hardware either, so it can be easily extended to the parameter identification field of other types of converters.

Key words

Capacitor / T-cascade modular multilevel converter (TMMC) / parameter identification / double-scale convolutional neural network

引用本文

代子阔, 张东升, 刘尚坤, 郭任, 陈丽娟, 汪诚. 基于双尺度卷积神经网络的TMMC直流母线电容参数辨识. 电源学报, 2026 , 24 (4) : 271 -283 . DOI: 10.13234/j.issn.2095-2805.2026.4.271

Zikuo DAI, Dongsheng ZHANG, Shangkun LIU, Ren GUO, Lijuan CHEN, Cheng WANG. Parameter Identification for DC-link Capacitors in TMMC Based on Double-scale Convolutional Neural Network[J]. Journal of Power Supply, 2026 , 24 (4) : 271 -283 . DOI: 10.13234/j.issn.2095-2805.2026.4.271

基金

收起

国网辽宁省电力有限公司科技资助项目(21YF-63)

参考文献

收起

文献

收起

参考文献引证文献

排序方式：

[1]

许铭林, 陈艳峰, 廖慧, 等. 轨道交通大功率模块化不间断供电电源可靠性和经济性分析[J]. 电源学报, 2021, 19(4): 162-171. Xu Minglin, Chen Yanfeng, Liao Hui, et al. Reliability and economic analysis of high-power modular UPS in urban rail transit applications[J]. Journal of Power Supply, 2021, 19(4): 162-171. (in Chinese)

[2]

傅军栋, 陈浩杰, 孙翔, 等. 基于EKF-Markov的UPS荷电状态预测与健康管理系统[J]. 电源学报, 2024, 22(6): 225-233. Fu Jundong, Chen Haojie, Sun Xiang, et al. SOC prediction and health management system of UPS based on EKF-Markov[J]. Journal of Power Supply, 2024, 22(6): 225-233. (in Chinese)

[3]

Wang Cheng, Hu Shaoqiang, Peng He, et al. Modulation coordinated voltage balance strategy for a dual-T-type modular multilevel converter[J]. IEEE Journal of Emerging and Selected Topics in Power Electronics, 2022, 10(6): 6648-6660.

[4]

Wang Cheng, Yang Hongkang, Hua Hongting, et al. Operation of dual-T-type modular multilevel converter for uninterrupted power supply under bridge failures[J]. IEEE Transactions on Industrial Electronics, 2023, 70(12): 11876-11886.

[5]

辛熙锴, 马柯. 一种MMC子模块电容ESR在线监测方法[J]. 电力电子技术, 2021, 55(7): 22-25. Xin Xikai, Ma Ke. An online ESR monitoring method for submodule capacitors in MMC[J]. Power Electronics, 2021, 55(7): 22-25. (in Chinese)

[6]

赵乐, 陈芳, 景旭川, 等. 外加注入信号的MMC子模块电压波动抑制策略[J]. 电力电子技术, 2021, 55(9): 128-130, 138. Zhao Le, Chen Fang, Jing Xuchuan, et al. Voltage fluctuation suppression strategy of MMC sub-module with additional injection signal[J]. Power Electronics, 2021, 55(9): 128-130, 138. (in Chinese)

[7]

Sundararajan P, Sathik M H M, Sasongko F, et al. Online condition monitoring system for DC-link capacitor in industrial power converters[J]. IEEE Transactions on Industry Applications, 2018, 54(5): 4775-4785.

[8]

毛向德, 王庆贤, 董唯光, 等. 小波包神经网络与数据降维的移相全桥变换器的故障诊断[J]. 电源学报, 2014, 12(4): 68-75. Mao Xiangde, Wang Qingxian, Dong Weiguang, et al. Phase-shift full bridge converter fault diagnosis based on wavelet packet and neural network and data dimensionnality reduction[J]. Journal of Power Supply, 2014, 12(4): 68-75. (in Chinese)

[9]

Lin Bing hao, Tsai J T, Lian K L. A non-invasive method for estimating circuit and control parameters of voltage source converters[J]. IEEE Transactions on Circuits and Systems I: Regular Papers, 2019, 66(12): 4911-4921.

[10]

Soliman H, Davari P, Wang Huai, et al. Capacitance estimation algorithm based on DC-link voltage harmonics using artificial neural network in three-phase motor drive systems[C]//2017 IEEE Energy Conversion Congress and Exposition (ECCE). Cincinnati, OH, USA, 2017: 5795-5802.

[11]

Yao Fang, Dong Chaoqun, Tang Shengxue, et al. Parameter identification of DC-link capacitor for electric vehicle based on IGWO-BP neural network[J]. IEEJ Transactions on Electrical and Electronic Engineering, 2021, 16(6): 861-870.

[12]

朱燕芳, 闫磊, 常康, 等. 基于深度卷积神经网络的电力系统故障预测[J]. 电源学报, 2024, 22(S1): 179-185. Zhu Yanfang, Yan Lei, Chang Kang, et al. Power system fault prediction based on deep convolutional neural network[J]. Journal of Power Supply, 2024, 22(S1): 179-185. (in Chinese)

[13]

毛明波, 孟昭亮, 高勇, 等. 基于PSO-BP神经网络的SiC MOSFET模块寿命预测方法研究与实现[J]. 电源学报, 2025, 23(1): 229-235, 258. Mao Mingbo, Meng Zhaoliang, Gao Yong, et al. Research and implementation of life prediction method for SiC MOSFET module based on PSO-BP neural network[J]. Journal of Power Supply, 2025, 23(1): 229-235, 258. (in Chinese)

[14]

赵小强, 罗维兰. 改进卷积Lenet-5神经网络的轴承故障诊断方法[J]. 电子测量与仪器学报, 2022, 36(6): 113-125. Zhao Xiaoqiang, Luo Weilan. Improved convolutional Lenet-5 neural network for bearing fault diagnosis[J]. Journal of Electronic Measurement and Instrumentation, 2022, 36(6): 113-125. (in Chinese)

[15]

崔江, 郭瑞东, 张卓然, 等. 基于改进DBN的发电机旋转整流器故障特征提取技术[J]. 中国电机工程学报, 2020, 40(7): 2369-2376. Cui Jiang, Guo Ruidong, Zhang Zhuoran, et al. Generator rotating rectifier fault feature extraction technique based on improved DBN[J]. Proceedings of the CSEE, 2020, 40(7): 2369-2376. (in Chinese)

2026年第24卷第4期

PDF下载

引用本文

BibTeX

文章信息

doi: 10.13234/j.issn.2095-2805.2026.4.271

接收时间：2023-11-09
首发时间：2026-10-29
出版时间：2026-04-30

补充材料

相关文章

文章信息

作者

出版历史

收稿日期：2023-11-09
修回日期：2024-02-03
录用日期：2024-02-21

基金

Science and Technology Program of State Grid Liaoning Electric Power Co., Ltd.(21YF-63)

国网辽宁省电力有限公司科技资助项目(21YF-63)

作者信息

^1.国网辽宁省电力有限公司，沈阳　110000

^2.国网辽宁省电力有限公司鞍山供电公司，鞍山　114000

^3.国网阜新供电公司，阜新　123100

^4.国网沈阳供电公司，沈阳　110003

^5.国网本溪供电公司，本溪　117020

^6.南京理工大学自动化学院，南京　210094

通讯作者:

汪诚（1988—　），男，通信作者，博士，副教授。研究方向：模块化电力电子变换器，主电路拓扑与性能控制。E-mail：chw714@njust.edu.cn。

参考文献

分享链接

https://castjournals.cast.org.cn/joweb/dyxb/CN/10.13234/j.issn.2095-2805.2026.4.271

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

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

BibTeX
EndNote
RefWorks
TxT