非负约束自动编码器在电缆早期故障识别中的应用
|
摘要
电缆早期故障的准确识别有助于降低电力系统的故障停电率和提高供电可靠性。在传统模式识别方法中,利于分类识别的有效特征通常难以选择,从而影响识别的准确度。鉴于此,将非负约束自动编码器(Non-negative Constrain Autoencoder, NCAE)堆叠形成的深度学习(Deep learning, DL)网络应用于电缆早期故障识别中。为了提高DL网络的学习效率,首先对故障相电流进行平稳小波变换,提取出一些具有相关性、冗余性的统计量、能量熵和信息熵等作为初级特征,其次堆叠多个NCAE构建出DL网络,通过预训练和微调机制,从初级特征中获得更易于早期故障分类识别的有效特征,最后利用Softmax分类器从正常状态和其他扰动信号中识别出早期故障。利用电缆电流仿真数据进行实验,结果表明与传统模式识别方法相比,所提方法识别准确率更高。
Accurate identification of cable incipient fault is helpful to reduce the failure rate of power system and improve the reliability of power supply. In the traditional pattern recognition method, it is difficult to select the efficient features, which are beneficial to classification, therefore, it would affect the accuracy of recognition. In view of this, this paper applies the Deep Learning(DL) network stacked from multiple non-negative constraint autoencoders to recognize cable incipient fault. In order to improve the learning efficiency of DL network, firstly, stationary wavelet transform of the fault phase current is used to extract the primary characteristics with correlation and redundancy, e. g. some statistics, energy entropy and information entropy. And then DL network is constructed by stacking multiple nonnegative constraint autoencoders. After preliminary training and fine-tuning training process, some effective features for recognition are learned from primary features. Finally, Softmax classifier is used to identify the cable incipient fault from normal state and other disturbances. Experiments on cable current simulation data are done, and the results show that the method is more accurate than the traditional pattern recognition method.
Accurate identification of cable incipient fault is helpful to reduce the failure rate of power system and improve the reliability of power supply. In the traditional pattern recognition method, it is difficult to select the efficient features, which are beneficial to classification, therefore, it would affect the accuracy of recognition. In view of this, this paper applies the Deep Learning(DL) network stacked from multiple non-negative constraint autoencoders to recognize cable incipient fault. In order to improve the learning efficiency of DL network, firstly, stationary wavelet transform of the fault phase current is used to extract the primary characteristics with correlation and redundancy, e. g. some statistics, energy entropy and information entropy. And then DL network is constructed by stacking multiple nonnegative constraint autoencoders. After preliminary training and fine-tuning training process, some effective features for recognition are learned from primary features. Finally, Softmax classifier is used to identify the cable incipient fault from normal state and other disturbances. Experiments on cable current simulation data are done, and the results show that the method is more accurate than the traditional pattern recognition method.
引文
[1]KULKARNI S,SANTOSO S,SHORT T A.Incipient fault location algorithm for underground cables[J].IEEETransaction on Smart Grid,2014,5(3):1165-1174.
[2]孙仲民,何正友,戴铭.基于灰色预测的电力电缆老化过程预警仿真研究[J].电力系统保护与控制,2012,40(13):55-60.SUN Zhongmin,HE Zhengyou,DAI Ming.Simulation study on pre-alarm of power cable aging process based on grey prediction[J].Power System Protection and Control,2012,40(13):55-60.
[3]SIDHU T S,XU Zhihan.Detection of incipient faults in distribution underground cables[J].IEEE Transactions on Power Delivery,2010,25(3):1363-1371.
[4]ZHANG Wenhai,XIAO Xiaoyong,ZHOU Kai,et al.Multi-cycle incipient fault detection and location for medium voltage underground cable[J].IEEE Transactions on Power Delivery,2017,32(3):1450-1459.
[5]ZHANG Chao,KANG Xiaoning,MA Xiuda,et al.On-line incipient faults detection in underground cables based on single-end sheath currents[C]//IEEE PES-Pacific Power and Energy Engineering Conference,October25-28,2016,Xi’an,China:795-799.
[6]戴铭.10 k V地下电缆早期故障检测与识别方法探讨[D].成都:西南交通大学,2012.DAI Ming.Discussion methods of incipient fault detection and identification in 10 kV underground cables[D].Chengdu:Southwest Jiaotong University,2012.
[7]SIDHU T S,XU Zhihan.Detection of incipient faults in distribution underground cables[J].IEEE Transactions on Power Delivery,2010,25(3):1363-1371.
[8]姚海燕,张静,留毅,等.基于多尺度小波判据和时频特征关联的电缆早期故障检测和识别方法[J].电力系统保护与控制,2015,43(9):115-123.YAO Haiyan,ZHANG Jing,LIU Yi,et al.A method of Incipient cable fault detection and identification based on multi-scale wavelet criterion and time-frequency feature correlation[J].Power System Protection and Control,2015,43(9):115-123.
[9]GHANBARI T.Kalman filter based incipient fault detection method for underground cables[J].IET Generation,Transmission&Distribution,2015,9(14):1988-1997.
[10]杨涛,黄军凯,许逵,等.基于深度学习的变压器故障诊断方法研究[J].电力大数据,2018,21(6):23-30.YANG Tao,HUANG Junkai,XU Kui,et al.Diagnosis method of power transformer fault based on deep learning[J].Power system and Big Data,2018,21(6):23-30.
[11]姜有泉,黄良,王波,等.基于DGA和深度置信网络的变压器内部故障诊断[J].武汉大学学报(工学版),2017,50(5):749-753.JIANG Youquan,HUANG Liang,WANG Bo,et al.Transformer internal fault diagnosis based on DGA and deep belief network[J].Engineering Journal of Wuhan University,2017,50(5):749-753.
[12]雷亚国,贾峰,周昕,等.基于深度学习理论的机械装备大数据健康监测方法[J].机械工程学报,2015,51(21):49-56.LEI Yaguo,JIA Feng,ZHOU Xin,et al.A deep learning-based method for machinery health monitoring with big data[J].Journal of Mechanical Engineering,2015,51(21):49-56.
[13]李智,侯兴哲,刘永相,等.基于深度学习的充电站容量规划方法[J].电力系统保护与控制,2017,45(21):67-73.LI Zhi,HOU Xingzhe,LIU Yongxiang,et al.A capacity planning method of charging station based on depth learning[J].Power System Protection and Control,2017,45(21):67-73.
[14]陈伟,何家欢,裴喜平.基于相空间重构和卷积神经网络的电能质量扰动分类[J].电力系统保护与控制,2018,46(14):87-93.CHEN Wei,HE Jiahuan,PEI Xiping.Classification for power quality disturbance based on phase-space reconstruction and convolution neural network[J].Power System Protection and Control,2018,46(14):87-93.
[15]HOSSEINI-AS E,ZURADA J M,NASRAOUI O.Deep learning of part-based representation of data using sparse autoencoders with nonnegativity constraints[J].IEEETransactions on Neural Networks and Learning Systems,2016,27(12):2486-2498.
[16]AYINDE B O,ZURADA J M.Deep learning of constrained autoencoders for enhanced understanding of data[J].IEEE Transactions on Neural Networks and Learning Systems,2018,29(9):3969-3979.
[17]NASON G P,SILVERMAN B W.The stationary wavelet transform and some statistical applications[J].Wavelets and Statistics,1995,103:281-299.
[18]COVER T M,HART P E.Nearest neighbor pattern classification[J].IEEE Transactions on Information Theory,1967,13(1):21-27.
[19]HEARST M A,DUMAIS S T,OSUNA E,et al.Support vector machines[J].IEEE Intelligent Systems and their Applications,1998,13(4):18-28.
[20]BREIMAN L.Random forests[J].Machine Learning,2001,45(1):5-32.
[21]BESSEC M,FOUQUAN J.Short-run electricity load forecasting with combinations of stationary wavelet transforms[J].European Journal of Operational Research,2018,264(1):149-164.
[22]SHARMA L D,SUNKARIA R K.Stationary wavelet transform based technique for automated external defibrillator using optimally selected classifiers[J].Measurement,2018,125:29-36.
[2]孙仲民,何正友,戴铭.基于灰色预测的电力电缆老化过程预警仿真研究[J].电力系统保护与控制,2012,40(13):55-60.SUN Zhongmin,HE Zhengyou,DAI Ming.Simulation study on pre-alarm of power cable aging process based on grey prediction[J].Power System Protection and Control,2012,40(13):55-60.
[3]SIDHU T S,XU Zhihan.Detection of incipient faults in distribution underground cables[J].IEEE Transactions on Power Delivery,2010,25(3):1363-1371.
[4]ZHANG Wenhai,XIAO Xiaoyong,ZHOU Kai,et al.Multi-cycle incipient fault detection and location for medium voltage underground cable[J].IEEE Transactions on Power Delivery,2017,32(3):1450-1459.
[5]ZHANG Chao,KANG Xiaoning,MA Xiuda,et al.On-line incipient faults detection in underground cables based on single-end sheath currents[C]//IEEE PES-Pacific Power and Energy Engineering Conference,October25-28,2016,Xi’an,China:795-799.
[6]戴铭.10 k V地下电缆早期故障检测与识别方法探讨[D].成都:西南交通大学,2012.DAI Ming.Discussion methods of incipient fault detection and identification in 10 kV underground cables[D].Chengdu:Southwest Jiaotong University,2012.
[7]SIDHU T S,XU Zhihan.Detection of incipient faults in distribution underground cables[J].IEEE Transactions on Power Delivery,2010,25(3):1363-1371.
[8]姚海燕,张静,留毅,等.基于多尺度小波判据和时频特征关联的电缆早期故障检测和识别方法[J].电力系统保护与控制,2015,43(9):115-123.YAO Haiyan,ZHANG Jing,LIU Yi,et al.A method of Incipient cable fault detection and identification based on multi-scale wavelet criterion and time-frequency feature correlation[J].Power System Protection and Control,2015,43(9):115-123.
[9]GHANBARI T.Kalman filter based incipient fault detection method for underground cables[J].IET Generation,Transmission&Distribution,2015,9(14):1988-1997.
[10]杨涛,黄军凯,许逵,等.基于深度学习的变压器故障诊断方法研究[J].电力大数据,2018,21(6):23-30.YANG Tao,HUANG Junkai,XU Kui,et al.Diagnosis method of power transformer fault based on deep learning[J].Power system and Big Data,2018,21(6):23-30.
[11]姜有泉,黄良,王波,等.基于DGA和深度置信网络的变压器内部故障诊断[J].武汉大学学报(工学版),2017,50(5):749-753.JIANG Youquan,HUANG Liang,WANG Bo,et al.Transformer internal fault diagnosis based on DGA and deep belief network[J].Engineering Journal of Wuhan University,2017,50(5):749-753.
[12]雷亚国,贾峰,周昕,等.基于深度学习理论的机械装备大数据健康监测方法[J].机械工程学报,2015,51(21):49-56.LEI Yaguo,JIA Feng,ZHOU Xin,et al.A deep learning-based method for machinery health monitoring with big data[J].Journal of Mechanical Engineering,2015,51(21):49-56.
[13]李智,侯兴哲,刘永相,等.基于深度学习的充电站容量规划方法[J].电力系统保护与控制,2017,45(21):67-73.LI Zhi,HOU Xingzhe,LIU Yongxiang,et al.A capacity planning method of charging station based on depth learning[J].Power System Protection and Control,2017,45(21):67-73.
[14]陈伟,何家欢,裴喜平.基于相空间重构和卷积神经网络的电能质量扰动分类[J].电力系统保护与控制,2018,46(14):87-93.CHEN Wei,HE Jiahuan,PEI Xiping.Classification for power quality disturbance based on phase-space reconstruction and convolution neural network[J].Power System Protection and Control,2018,46(14):87-93.
[15]HOSSEINI-AS E,ZURADA J M,NASRAOUI O.Deep learning of part-based representation of data using sparse autoencoders with nonnegativity constraints[J].IEEETransactions on Neural Networks and Learning Systems,2016,27(12):2486-2498.
[16]AYINDE B O,ZURADA J M.Deep learning of constrained autoencoders for enhanced understanding of data[J].IEEE Transactions on Neural Networks and Learning Systems,2018,29(9):3969-3979.
[17]NASON G P,SILVERMAN B W.The stationary wavelet transform and some statistical applications[J].Wavelets and Statistics,1995,103:281-299.
[18]COVER T M,HART P E.Nearest neighbor pattern classification[J].IEEE Transactions on Information Theory,1967,13(1):21-27.
[19]HEARST M A,DUMAIS S T,OSUNA E,et al.Support vector machines[J].IEEE Intelligent Systems and their Applications,1998,13(4):18-28.
[20]BREIMAN L.Random forests[J].Machine Learning,2001,45(1):5-32.
[21]BESSEC M,FOUQUAN J.Short-run electricity load forecasting with combinations of stationary wavelet transforms[J].European Journal of Operational Research,2018,264(1):149-164.
[22]SHARMA L D,SUNKARIA R K.Stationary wavelet transform based technique for automated external defibrillator using optimally selected classifiers[J].Measurement,2018,125:29-36.