电力电子装置的智能故障诊断
|
摘要
随着电力电子技术的高速发展,新的电力电子器件的不断涌现,电力电子技术的应用日益深入到工业生产和社会生活的各个方面。电力电子设备的广泛应用使得对设备的可维护性要求越来越突出。高电压、大电流的电力电子装置的功率器件较多,发生故障时用常规的检测方法费时费力,针对故障本身的特点,人工智能、神经网络、小波分析等新兴的学科已成功应用于电力电子设备的故障诊断。智能化的故障自动诊断方法,有利于快速分析确定故障的部位和性质,缩短电力电子电路的运行停机时间,提高效率,减少损失。
故障特征提取和识别方法的研究对发展和完善电力电子装置的智能故障诊断技术有着重要的作用。应用小波包能量法提取出电力电子装置在各种状态下电压及电流信号的能量特征向量,并将它们进行数据融合作为神经元网络故障分类器的输入向量,由神经元网络故障分类器对各种故障进行识别和诊断。以电力电子整流装置主电路故障为例进行了仿真实验,试验结果表明该方法无需数学模型就能快速准确的完成故障定位诊断。
利用小波及小波包分析﹑神经网络及系统辨识等理论实现了三电平逆变装置的故障诊断软件系统设计,诊断系统是直接利用MATLAB自身携带的GUIDE(图形用户界面开发环境)完成了可视化设计,实现了界面与功能的一体化设计。经试验验证该故障诊断系统故障诊断率高﹑诊断速度快﹑操作简单﹑实用性强。
With the rapid development of power electronics and continual emerging of novel power electronic devices, power electronic equipments have been increasingly applied to all aspects of industry and living. The requirement of its maintainability is also more and more important. Because there are many devices in these circuits of high voltage, strong current systems, one will take much time and make many efforts to deal with it using normal fault diagnosis methods. Aiming at the own characteristics of fault, new technologies such as artificial intelligence, neural network, wavelet analysis, etc have been successfully used to fault diagnosis of power electronics. Intellectualized automatic fault diagnosis method has the advantage of rapid analyzing and localizing the fault, cutting the stop time, increasing the efficiency and reducing the loss.
The research on failure feature extraction and failure to identify plays an important role in developing and improving the intelligent fault diagnostic. These signal feature vectors of various states about voltage and current of power electronic device, which are extracted by using the wavelet packet energy method, are integrated as the input vectors of the neural network failure classifier, identified and diagnosed various faults by the neural network classifier. The result of the simulate experiment with the main circuit of electronic power electronic rectifier device failure as an example, shows that the method can deal with the location of fault diagnosis rapidly and accurately without mathematical model.
The software system design of fault diagnosis of three-level inverter device are realized by using these theories of wavelet and wavelet packet analysis, neural network, system identification and so on. The Visual diagnostic system is designed by direct using the GUIDE (graphical user interface development environment) of MATLAB. It has realized one design of the Interface and function. The tests results show that the fault diagnosis system have many advantages of high rate of fault diagnosis, diagnosis rapidly, operate simply and strong practical.
故障特征提取和识别方法的研究对发展和完善电力电子装置的智能故障诊断技术有着重要的作用。应用小波包能量法提取出电力电子装置在各种状态下电压及电流信号的能量特征向量,并将它们进行数据融合作为神经元网络故障分类器的输入向量,由神经元网络故障分类器对各种故障进行识别和诊断。以电力电子整流装置主电路故障为例进行了仿真实验,试验结果表明该方法无需数学模型就能快速准确的完成故障定位诊断。
利用小波及小波包分析﹑神经网络及系统辨识等理论实现了三电平逆变装置的故障诊断软件系统设计,诊断系统是直接利用MATLAB自身携带的GUIDE(图形用户界面开发环境)完成了可视化设计,实现了界面与功能的一体化设计。经试验验证该故障诊断系统故障诊断率高﹑诊断速度快﹑操作简单﹑实用性强。
With the rapid development of power electronics and continual emerging of novel power electronic devices, power electronic equipments have been increasingly applied to all aspects of industry and living. The requirement of its maintainability is also more and more important. Because there are many devices in these circuits of high voltage, strong current systems, one will take much time and make many efforts to deal with it using normal fault diagnosis methods. Aiming at the own characteristics of fault, new technologies such as artificial intelligence, neural network, wavelet analysis, etc have been successfully used to fault diagnosis of power electronics. Intellectualized automatic fault diagnosis method has the advantage of rapid analyzing and localizing the fault, cutting the stop time, increasing the efficiency and reducing the loss.
The research on failure feature extraction and failure to identify plays an important role in developing and improving the intelligent fault diagnostic. These signal feature vectors of various states about voltage and current of power electronic device, which are extracted by using the wavelet packet energy method, are integrated as the input vectors of the neural network failure classifier, identified and diagnosed various faults by the neural network classifier. The result of the simulate experiment with the main circuit of electronic power electronic rectifier device failure as an example, shows that the method can deal with the location of fault diagnosis rapidly and accurately without mathematical model.
The software system design of fault diagnosis of three-level inverter device are realized by using these theories of wavelet and wavelet packet analysis, neural network, system identification and so on. The Visual diagnostic system is designed by direct using the GUIDE (graphical user interface development environment) of MATLAB. It has realized one design of the Interface and function. The tests results show that the fault diagnosis system have many advantages of high rate of fault diagnosis, diagnosis rapidly, operate simply and strong practical.
引文
[1]王锋,华胜,万浩平.我国电力电子技术发展展望[J].科技广场,2008.8.
[2]郑家伟.新能源发电系统中的电力电子变换[J].控制理论与应用,2008,25(1):125-132.
[3] Bimal K Bose. Energy, Environment, and Advances in Power Electronics [J].IEEE Trans on Power Electronics, 2000, 15(4): 688-701.
[4] Sang Bin Lee, Rangarajan M. Tallam, and Thomas G. Habetler.A Robust, On-Line Turn Fault Detection Technique for Induction Machines Based on Monitoring the Sequence Component Impedance Matrix. IEEE Transactions on Power Electronics, Vol.18, No.3, May2003:865~872.
[5]李微,谭阳红,彭永进.基于小波神经网络的电力电子电路故障模式识别[J].继电器,2005: 33(14):82-86.
[6]李敏远,陈如清.一种基于模式识别的可控整流电路故障诊断方法[J].电工技术学报,2004,19(7):53-58.
[7]余志红.电气设备在线监测与故障诊断技术的现状与前景[J].江西电力,2008,32(5):28-31.
[8]杨寅哲.智能故障诊断标准化技术研究[J].科技创新导报,2008,21:156-158.
[9]张维勇,程凡,魏臻.一种综合智能故障诊断方法的研究[J].合肥工业大学学报(自然科学版),2005,28(9):1104-1108.
[10]黄念慈,黄山.旋转整流器故障检测的一种方法[J].电工技术学报,1996,11(4):50-53.
[11]张萍,王贵增,周东华.动态系统的故障诊断方法[J].控制理论与应用,2000, 17(2):153~157.
[12] Matthew Senesky,Gabriel Eirea, T.John Koo, Hybrid Modeling and Control of Power Electronics, Springer-verlag, HSCC2003, LNCS2623, 2003:450-465.
[13]梁中华,田茂芹,任敏.基于人工神经网络的电力电子电路故障诊断[J].沈阳工业大学学报,2004,26(2):140-142.
[14] Daubechies I.Ten Lectures on Wavelets[M].CBMS-NSF Series in Appl Math, Philadelphia:SIAM,1991,58-59.
[15] [美]崔锦泰著,程正兴译.小波分析导论[M].西安交通大学出版社,1995.
[16]杨福生.小波变换的工程分析与应用[M].北京:科学出版社,2003.
[17]飞思科技产品研发中心编著. MATLAB6.5辅助小波分析与应用[M].北京:电子工业出版社,2003,1.
[18]黄艳芳,何福顺,杨洁明.基于小波包预处理的神经网络故障模式识别的研究[J].煤矿机械,2002,5:18~20.
[19] Tae-Jim Kim,Ju-Won Baek, Jim-Hong Jeon, Geun-Hie Rim, Cheul-U Kim. A diagnosis method of DC/DC converter aging based on the variation of parasitic. Proceedings of the 30th annual conference of the IEEE Industrial Electronics Society.2004.3037-3041.
[20]董俊宏,王瑛剑,李小珉.集成运放放大电路的噪声分析[J].电气电子教学学报,2006,28(2):29-31.
[21]让光林.相互作用量子场论的白噪声方法[D].华中科技大学,2003.
[22]周东华,叶银忠著.现代故障处理与容错控制[M].清华大学出版社,2000,6.
[23]徐德鸿,马皓编著.电力电子装置故障自动诊断[M].北京:科学出版社,2001,3.
[24] Berezansky Yu M, Lytvynov E, Mierzejewski D A. The Jocobi field of a process. Ukrainian Math.J., 2003, 55:853-858.
[25]魏明,朱贤斌,季仲梅. WCDMA系统中噪声滤波器的设计与实现. [J].电声器件与电路, 2008,32(5):29-31.
[26]王欣,王德隽.离散信号的滤波[M].北京:电子工业出版社,2002.
[27] KIM Chang-joo, LEE Hyuck-jae. Adaptive acquisition of PN sequences for DSSS communications[J].IEEE Trans. on Commun.,1998,46(8):993-996.
[28]熊小龙,王洁,由大德等.基于数据融合的智能故障诊断仿真研究[J].现代防御技术, 2008,36(5):161-164.
[29]何友,王国宏,陆大金,等.多传感器信息融合及应用[M].北京:电子工业出版社, 2000.
[30]蓝金辉,马宝华,蓝天,等. D-S证据理论数据融合方法在目标识别中的应用[J].清华大学学报(自然科学版), 2001, 41(2): 53-57.
[31]焦竹青,熊伟丽,张林,徐保国.基于信任度的多传感器数据融合及其应用[J].东南大学学报(自然科学版), 2008, 38(1): 253-257.
[32] Moon Hyeong-Soon,Kim Yong Back,Beattid Robert J.Multi-sensor data fusion for improving performance and reliability of fully automatic welding system[J]. International Journal of Advance Manufacture Technology,2006,28(3-4):286-293.
[33]阎平凡,张长水编著.人工神经网络与模拟进化计算[J].北京:清华大学出版社,2000,11.
[34]崔容一,洪炳熔.关于三层前馈神经网络隐层构建问题的研究.计算机研究与应用,2004,41(4):524-530.
[35]吴成茂,范九伦.确定RBF神经网络节点数的最大矩阵元法.计算机工程与应用,2004,20:77~79.
[36] Bo Li, Mo-Yuen Chow, Yodyium Tipsuwan, and James C. Hung. Neural-Network-Based Motor Rolling Bearing Fault Diagnosis. IEEE Transactions on Industrial Electronics, Vol. 47,No. 5, October 2000:383~389.
[37] Mohamed A. Awadallah,and Medhat M. Morcos.Application of AI Tools in Fault Diagnosis of Electrical Machines and Drives—An Overview. IEEE Transactions on Energy Conversion, Vol. 18, No. 2, June 2003:245~251.
[38] J. M. Cardoso and A. M. S. Mendes.Converter fault diagnosis in variable speed DC drives by Park’s vector approach.in Proc. IEEE Int.Symp. Ind. Electron., 1997:497~500.
[39]李作泳,蔡辉,丁晶.BP网络的学习能力与推广能力之间满足的不确定关系式[J].四川大学学报(工程科学版),2003,35(l):16-19.
[40] G.-B.Huang.Learning capability and storage capacity of two-hidden-layer feedforward networks[J].IEEE Trans.Neural Networks,2003,14(6):274-281.
[41] G.-B.Huang.Real-time Learning Capability of Neural Networks[J].IEEE Trans.On NN,2004.
[42]张兆宁.小波分析、模糊理论及神经网络在电力系统综合自动化中的应用研究.[上海交通大学博士后工作报告],2002.3
[43] Jee-Shing Wang,C.S.George Lee.Self-Adaptive Recurrent Neural-Fuzzy Control of an Autonomous Underwater Vehicle.IEEE Trans.on Robotics and Automation, 2003, Vol.19,No.2,pp.283-294
[44]田佳禾,孙士慧,赵仕俊.基于小波神经网络的故障诊断方法研究[J].计算机系统应用, 2008,36(5):43-46.
[45] Abhinav Saxena, Ashraf Saad. Evolving an artificial neural network classifier for condition monitoring of ro-tating mechanical systems. Applied Soft Computing, 2007,7(7):441-454.
[46]王宏刚,钱峰.基于遗传算法的前向神经网络结构优化[J].控制工程,2007,(4):387-390.
[47] B.Samanta. Artificial neural networks and genetic al-gorithms for gear fault detection, Mech. Syst. Signal Process, 2004,18(5):1273-1282.
[48]李显宏.MATLAB7.x界面设计与编程技巧[M].北京:电子工业出版社,2006.12.
[49] MATLAB Help Documentation.MathWorks 7.3.
[50]王俊国,王永翼,万淑芸.MATLAB图形用户界面环境下的实时控制系统[J].仪表技术与传感器,2004,(9):27~29.
[51]苏金明,刘宏,刘波. MATLAB高级编程[M].北京:电子工业出版社, 2005.
[52]苏金明,黄国明,刘波. MATLAB与外部程序接口[M].北京:电子工业出版社, 2004.
[53] MATLAB Users' Guide. MathWorks Inc[EB/OL] . 2004.
[54]施小红,周佳.精通GUI图形界面编程[M].北京:北京大学出版社, 2003
[55]李亚军,赵刚,王华,等.在MATLAB图形用户界面设计中使用ActiveX控件.2006 ,34(12):135-137.
[56]王默玉,宗伟,等.基于MATLAB的图形用户界面的构造方式与应用[J].现代电力,2002,19(1):76-82.
[57]夏文芳,吴鸿修,李享元.基于USB的实时数据采集系统及其在MATLAB中的应用[J].计算机应用,2003,2:15-17.
[2]郑家伟.新能源发电系统中的电力电子变换[J].控制理论与应用,2008,25(1):125-132.
[3] Bimal K Bose. Energy, Environment, and Advances in Power Electronics [J].IEEE Trans on Power Electronics, 2000, 15(4): 688-701.
[4] Sang Bin Lee, Rangarajan M. Tallam, and Thomas G. Habetler.A Robust, On-Line Turn Fault Detection Technique for Induction Machines Based on Monitoring the Sequence Component Impedance Matrix. IEEE Transactions on Power Electronics, Vol.18, No.3, May2003:865~872.
[5]李微,谭阳红,彭永进.基于小波神经网络的电力电子电路故障模式识别[J].继电器,2005: 33(14):82-86.
[6]李敏远,陈如清.一种基于模式识别的可控整流电路故障诊断方法[J].电工技术学报,2004,19(7):53-58.
[7]余志红.电气设备在线监测与故障诊断技术的现状与前景[J].江西电力,2008,32(5):28-31.
[8]杨寅哲.智能故障诊断标准化技术研究[J].科技创新导报,2008,21:156-158.
[9]张维勇,程凡,魏臻.一种综合智能故障诊断方法的研究[J].合肥工业大学学报(自然科学版),2005,28(9):1104-1108.
[10]黄念慈,黄山.旋转整流器故障检测的一种方法[J].电工技术学报,1996,11(4):50-53.
[11]张萍,王贵增,周东华.动态系统的故障诊断方法[J].控制理论与应用,2000, 17(2):153~157.
[12] Matthew Senesky,Gabriel Eirea, T.John Koo, Hybrid Modeling and Control of Power Electronics, Springer-verlag, HSCC2003, LNCS2623, 2003:450-465.
[13]梁中华,田茂芹,任敏.基于人工神经网络的电力电子电路故障诊断[J].沈阳工业大学学报,2004,26(2):140-142.
[14] Daubechies I.Ten Lectures on Wavelets[M].CBMS-NSF Series in Appl Math, Philadelphia:SIAM,1991,58-59.
[15] [美]崔锦泰著,程正兴译.小波分析导论[M].西安交通大学出版社,1995.
[16]杨福生.小波变换的工程分析与应用[M].北京:科学出版社,2003.
[17]飞思科技产品研发中心编著. MATLAB6.5辅助小波分析与应用[M].北京:电子工业出版社,2003,1.
[18]黄艳芳,何福顺,杨洁明.基于小波包预处理的神经网络故障模式识别的研究[J].煤矿机械,2002,5:18~20.
[19] Tae-Jim Kim,Ju-Won Baek, Jim-Hong Jeon, Geun-Hie Rim, Cheul-U Kim. A diagnosis method of DC/DC converter aging based on the variation of parasitic. Proceedings of the 30th annual conference of the IEEE Industrial Electronics Society.2004.3037-3041.
[20]董俊宏,王瑛剑,李小珉.集成运放放大电路的噪声分析[J].电气电子教学学报,2006,28(2):29-31.
[21]让光林.相互作用量子场论的白噪声方法[D].华中科技大学,2003.
[22]周东华,叶银忠著.现代故障处理与容错控制[M].清华大学出版社,2000,6.
[23]徐德鸿,马皓编著.电力电子装置故障自动诊断[M].北京:科学出版社,2001,3.
[24] Berezansky Yu M, Lytvynov E, Mierzejewski D A. The Jocobi field of a process. Ukrainian Math.J., 2003, 55:853-858.
[25]魏明,朱贤斌,季仲梅. WCDMA系统中噪声滤波器的设计与实现. [J].电声器件与电路, 2008,32(5):29-31.
[26]王欣,王德隽.离散信号的滤波[M].北京:电子工业出版社,2002.
[27] KIM Chang-joo, LEE Hyuck-jae. Adaptive acquisition of PN sequences for DSSS communications[J].IEEE Trans. on Commun.,1998,46(8):993-996.
[28]熊小龙,王洁,由大德等.基于数据融合的智能故障诊断仿真研究[J].现代防御技术, 2008,36(5):161-164.
[29]何友,王国宏,陆大金,等.多传感器信息融合及应用[M].北京:电子工业出版社, 2000.
[30]蓝金辉,马宝华,蓝天,等. D-S证据理论数据融合方法在目标识别中的应用[J].清华大学学报(自然科学版), 2001, 41(2): 53-57.
[31]焦竹青,熊伟丽,张林,徐保国.基于信任度的多传感器数据融合及其应用[J].东南大学学报(自然科学版), 2008, 38(1): 253-257.
[32] Moon Hyeong-Soon,Kim Yong Back,Beattid Robert J.Multi-sensor data fusion for improving performance and reliability of fully automatic welding system[J]. International Journal of Advance Manufacture Technology,2006,28(3-4):286-293.
[33]阎平凡,张长水编著.人工神经网络与模拟进化计算[J].北京:清华大学出版社,2000,11.
[34]崔容一,洪炳熔.关于三层前馈神经网络隐层构建问题的研究.计算机研究与应用,2004,41(4):524-530.
[35]吴成茂,范九伦.确定RBF神经网络节点数的最大矩阵元法.计算机工程与应用,2004,20:77~79.
[36] Bo Li, Mo-Yuen Chow, Yodyium Tipsuwan, and James C. Hung. Neural-Network-Based Motor Rolling Bearing Fault Diagnosis. IEEE Transactions on Industrial Electronics, Vol. 47,No. 5, October 2000:383~389.
[37] Mohamed A. Awadallah,and Medhat M. Morcos.Application of AI Tools in Fault Diagnosis of Electrical Machines and Drives—An Overview. IEEE Transactions on Energy Conversion, Vol. 18, No. 2, June 2003:245~251.
[38] J. M. Cardoso and A. M. S. Mendes.Converter fault diagnosis in variable speed DC drives by Park’s vector approach.in Proc. IEEE Int.Symp. Ind. Electron., 1997:497~500.
[39]李作泳,蔡辉,丁晶.BP网络的学习能力与推广能力之间满足的不确定关系式[J].四川大学学报(工程科学版),2003,35(l):16-19.
[40] G.-B.Huang.Learning capability and storage capacity of two-hidden-layer feedforward networks[J].IEEE Trans.Neural Networks,2003,14(6):274-281.
[41] G.-B.Huang.Real-time Learning Capability of Neural Networks[J].IEEE Trans.On NN,2004.
[42]张兆宁.小波分析、模糊理论及神经网络在电力系统综合自动化中的应用研究.[上海交通大学博士后工作报告],2002.3
[43] Jee-Shing Wang,C.S.George Lee.Self-Adaptive Recurrent Neural-Fuzzy Control of an Autonomous Underwater Vehicle.IEEE Trans.on Robotics and Automation, 2003, Vol.19,No.2,pp.283-294
[44]田佳禾,孙士慧,赵仕俊.基于小波神经网络的故障诊断方法研究[J].计算机系统应用, 2008,36(5):43-46.
[45] Abhinav Saxena, Ashraf Saad. Evolving an artificial neural network classifier for condition monitoring of ro-tating mechanical systems. Applied Soft Computing, 2007,7(7):441-454.
[46]王宏刚,钱峰.基于遗传算法的前向神经网络结构优化[J].控制工程,2007,(4):387-390.
[47] B.Samanta. Artificial neural networks and genetic al-gorithms for gear fault detection, Mech. Syst. Signal Process, 2004,18(5):1273-1282.
[48]李显宏.MATLAB7.x界面设计与编程技巧[M].北京:电子工业出版社,2006.12.
[49] MATLAB Help Documentation.MathWorks 7.3.
[50]王俊国,王永翼,万淑芸.MATLAB图形用户界面环境下的实时控制系统[J].仪表技术与传感器,2004,(9):27~29.
[51]苏金明,刘宏,刘波. MATLAB高级编程[M].北京:电子工业出版社, 2005.
[52]苏金明,黄国明,刘波. MATLAB与外部程序接口[M].北京:电子工业出版社, 2004.
[53] MATLAB Users' Guide. MathWorks Inc[EB/OL] . 2004.
[54]施小红,周佳.精通GUI图形界面编程[M].北京:北京大学出版社, 2003
[55]李亚军,赵刚,王华,等.在MATLAB图形用户界面设计中使用ActiveX控件.2006 ,34(12):135-137.
[56]王默玉,宗伟,等.基于MATLAB的图形用户界面的构造方式与应用[J].现代电力,2002,19(1):76-82.
[57]夏文芳,吴鸿修,李享元.基于USB的实时数据采集系统及其在MATLAB中的应用[J].计算机应用,2003,2:15-17.