基于EMD的ICA降噪方法在电厂开关柜局部放电信号中的应用
|
摘要
电厂开关柜运行环境复杂,开关柜局部放电PD(partial discharge)信号常湮没在强烈的干扰噪声中,无法准确检测出可靠结果。为了提高电厂开关柜局部放电检测的可靠性,首先将检测到的含噪PD信号进行经验模态分解EMD(empirical mode decomposition),筛选出合适的本征模态分量IMF(intrinsic mode function)构造虚拟噪声通道;再将虚拟噪声通道与含噪PD信号作为输入,利用独立分量分析ICA(independent component analysis)算法将PD信号与噪声信号分离,提取PD信号。通过在仿真信号与实测信号中的应用表明,该方法能够较好地实现对含噪PD信号的噪声消除。与常见的小波降噪方法相比,该方法降噪效果更好,不仅降低了噪声的干扰,同时还很好地保持了原信号的特征。
The partial discharge(PD)signal of a switch cabinet in a power plant is often trapped in the strong interference noise due to the complicated operating environment,thus a reliable result is impossible to be detected accurately.To improve the reliability of detection,the detected noisy PD signal is first decomposed by empirical mode decomposition(EMD),and some suitable intrinsic mode function(IMF)components are selected to construct a virtual noise channel. Then,the virtual noise channel and the noisy PD signal are taken as input,and the independent component analysis(ICA)method is used to separate the PD signal from the noise signal. In this way,the PD signal can be extracted. The application of the proposed method in simulation and measured signals shows that it can achieve better noise cancellation for noisy PD signals. Compared with the common wavelet de-noise method,this method has better de-noise effect;it not only reduces noise interference,but also keeps the characteristics of the original signal well.
The partial discharge(PD)signal of a switch cabinet in a power plant is often trapped in the strong interference noise due to the complicated operating environment,thus a reliable result is impossible to be detected accurately.To improve the reliability of detection,the detected noisy PD signal is first decomposed by empirical mode decomposition(EMD),and some suitable intrinsic mode function(IMF)components are selected to construct a virtual noise channel. Then,the virtual noise channel and the noisy PD signal are taken as input,and the independent component analysis(ICA)method is used to separate the PD signal from the noise signal. In this way,the PD signal can be extracted. The application of the proposed method in simulation and measured signals shows that it can achieve better noise cancellation for noisy PD signals. Compared with the common wavelet de-noise method,this method has better de-noise effect;it not only reduces noise interference,but also keeps the characteristics of the original signal well.
引文
[1]潘长明,刘刚,熊炬,等(Pan Changming,Liu Gang,Xiong Ju,et al).高压开关柜绝缘事故的分析及防范措施(High voltage switchgear insulation accident’s analysis and counter measures)[J].高压电器(High Voltage Apparatus),2011,47(7):90-93.
[2]刘云鹏,王会斌,王娟,等(Liu Yunpeng,Wang Huibin,Wang Juan,et al).高压开关柜局部放电UHF在线监测系统的研究(Research on on-line partial discharge UHF monitoring system for high voltage switchboard)[J].高压电器(High Voltage Apparatus),2009,45(1):15-17.
[3]王浩鸣,唐翀,吴莉萍,等(Wang Haoming,Tang Chong,Wu Liping,et al).基于设备状态评价的配电网动态可靠性评估(Dynamic reliability assessment of distribution network based on equipment state evaluation model)[J].电力系统及其自动化学报(Proceedings of the CSU-EPSA),2017,29(7):68-74.
[4]余先川,胡丹.盲源分离理论与应用[M].北京:科学出版社,2011.
[5]赵阳,夏欢,邱晓晖,等(Zhao Yang,Xia Huan,Qiu Xiaohui,et al).基于独立分量分析算法的医疗电子设备辐射电磁干扰EMI噪声分析方法(The radiated EMI analysis method for electromedical equipments based on ICA)[J].电工技术学报(Transactions of China Electrotechnical Society),2015,30(17):180-190.
[6]叶林,刘鹏(Ye Lin,Liu Peng).基于经验模态分解和支持向量机的短期风电功率组合预测模型(Combined model based on EMD-SVM for short-term wind power prediction)[J].中国电机工程学报(Proceedings of the CSEE),2011,31(31):102-108.
[7]孙曙光,王秋萌,闫欢,等(Sun Shuguang,Wang Qiumeng,Yan Huan,et al). EEMD在电力谐波检测中的应用(Application of EEMD to power harmonic detection)[J].电力系统及其自动化学报(Proceedings of the CSUEPSA),2016,28(8):25-31.
[8]康守强,王玉静,杨广学,等(Kang Shouqiang,Wang Yujing,Yang Guangxue,et al).基于经验模态分解和超球多类支持向量机的滚动轴承故障诊断方法(Rolling bearing fault diagnosis method using empirical mode decomposition and hypersphere multiclass support vector machine)[J].中国电机工程学报(Proceedings of the CSEE),2011,31(14):96-102.
[9]马宏伟,张大伟,曹现刚,等(Ma Hongwei,Zhang Dawei,Cao Xiangang,et al).基于EMD的振动信号去噪方法研究(Vibration signal de-noising method based on empirical mode decomposition)[J].振动与冲击(Journal of Vibration and Shock),2016,35(22):38-40.
[10]胡爱军,孙敬敬,向玲(Hu Aijun,Sun Jingjing,Xiang Ling).经验模态分解中的模态混叠问题(Mode mixing in empirical mode decomposition)[J].振动.测试与诊断(Journal of Vibration,Measurement&Diagnosis),2011,31(4):429-434,532-533.
[11]李辉,李洋,杨东,等(Li Hui,Li Yang,Yang Dong,et al).基于EMD相关去噪的风电机组振动噪声抑制及特征频率提取(Noise suppression and characteristic frequency extraction of wind turbine vibration based on EMD correlation denoising)[J].电机与控制学报(Electric Machines and Control),2016,20(1):73-80.
[12]杨凯,张认成,杨建红,等(Yang Kai,Zhang Rencheng,Yang Jianhong,et al).基于频率约束独立分量分析的开关柜局部放电故障识别(Partial discharge fault identification of switchgear based on frequency constrained independent component analysis)[J].高电压技术(High Voltage Engineering),2014,40(11):3452-3460.
[13]赵熙,杨洪耕(Zhao Xi,Yang Honggeng).基于快速独立分量分析的系统侧谐波阻抗计算方法(Method of calculating system-side harmonic impedance based on FastICA)[J].电力系统自动化(Automation of Electric Power Systems),2015,39(23):139-144,151.
[14]谢庆,张丽君,程述一,等(Xie Qing,Zhang Lijun,Cheng Shuyi,et al).快速独立分量分析算法在局放超声阵列信号去噪中的应用(Application of the FastICA algorithm to PD ultrasonic array signal de-noising)[J].中国电机工程学报(Proceedings of the CSEE),2012,32(18):160-166.
[15]季忠,金涛,杨炯明,等(Ji Zhong,Jin Tao,Yang Jiongming,et al).虚拟噪声通道在基于ICA消噪过程中的应用(Application of virtual noise channels in the processing of signal de-noising with ICA)[J].中国机械工程(China Mechanical Engineering),2005,16(4):350-352,364.
[16]焦卫东(Jiao Weidong).基于独立分量分析的旋转机械故障诊断方法研究(Research on Methods for Fault Diagnosis of Rotating Machines Based on Independent Component Analysis)[D].杭州:浙江大学机械与能源工程学院(Hangzhou:College of Mechanical and Energy Engineering,Zhejiang University),2003.
[17]陈东方,吴先良(Chen Dongfang,Wu Xianliang).采用EMD方法消除瞬态散射回波中的高斯白噪声干扰(Recovery of signal from transient scattered response contaminated by Gaussian white noise based on EMD method)[J].电子学报(Acta Electronica Sinica),2004,32(3):496-498.
[18]钱勇,黄成军,陈陈,等(Qian Yong,Huang Chengjun,Chen Chen,et al).基于经验模态分解的局部放电去噪方法(Denoising of partial discharge based on empirical mode decomposition)[J].电力系统自动化(Automation of Electric Power Systems),2005,29(12):53-56,60.
[2]刘云鹏,王会斌,王娟,等(Liu Yunpeng,Wang Huibin,Wang Juan,et al).高压开关柜局部放电UHF在线监测系统的研究(Research on on-line partial discharge UHF monitoring system for high voltage switchboard)[J].高压电器(High Voltage Apparatus),2009,45(1):15-17.
[3]王浩鸣,唐翀,吴莉萍,等(Wang Haoming,Tang Chong,Wu Liping,et al).基于设备状态评价的配电网动态可靠性评估(Dynamic reliability assessment of distribution network based on equipment state evaluation model)[J].电力系统及其自动化学报(Proceedings of the CSU-EPSA),2017,29(7):68-74.
[4]余先川,胡丹.盲源分离理论与应用[M].北京:科学出版社,2011.
[5]赵阳,夏欢,邱晓晖,等(Zhao Yang,Xia Huan,Qiu Xiaohui,et al).基于独立分量分析算法的医疗电子设备辐射电磁干扰EMI噪声分析方法(The radiated EMI analysis method for electromedical equipments based on ICA)[J].电工技术学报(Transactions of China Electrotechnical Society),2015,30(17):180-190.
[6]叶林,刘鹏(Ye Lin,Liu Peng).基于经验模态分解和支持向量机的短期风电功率组合预测模型(Combined model based on EMD-SVM for short-term wind power prediction)[J].中国电机工程学报(Proceedings of the CSEE),2011,31(31):102-108.
[7]孙曙光,王秋萌,闫欢,等(Sun Shuguang,Wang Qiumeng,Yan Huan,et al). EEMD在电力谐波检测中的应用(Application of EEMD to power harmonic detection)[J].电力系统及其自动化学报(Proceedings of the CSUEPSA),2016,28(8):25-31.
[8]康守强,王玉静,杨广学,等(Kang Shouqiang,Wang Yujing,Yang Guangxue,et al).基于经验模态分解和超球多类支持向量机的滚动轴承故障诊断方法(Rolling bearing fault diagnosis method using empirical mode decomposition and hypersphere multiclass support vector machine)[J].中国电机工程学报(Proceedings of the CSEE),2011,31(14):96-102.
[9]马宏伟,张大伟,曹现刚,等(Ma Hongwei,Zhang Dawei,Cao Xiangang,et al).基于EMD的振动信号去噪方法研究(Vibration signal de-noising method based on empirical mode decomposition)[J].振动与冲击(Journal of Vibration and Shock),2016,35(22):38-40.
[10]胡爱军,孙敬敬,向玲(Hu Aijun,Sun Jingjing,Xiang Ling).经验模态分解中的模态混叠问题(Mode mixing in empirical mode decomposition)[J].振动.测试与诊断(Journal of Vibration,Measurement&Diagnosis),2011,31(4):429-434,532-533.
[11]李辉,李洋,杨东,等(Li Hui,Li Yang,Yang Dong,et al).基于EMD相关去噪的风电机组振动噪声抑制及特征频率提取(Noise suppression and characteristic frequency extraction of wind turbine vibration based on EMD correlation denoising)[J].电机与控制学报(Electric Machines and Control),2016,20(1):73-80.
[12]杨凯,张认成,杨建红,等(Yang Kai,Zhang Rencheng,Yang Jianhong,et al).基于频率约束独立分量分析的开关柜局部放电故障识别(Partial discharge fault identification of switchgear based on frequency constrained independent component analysis)[J].高电压技术(High Voltage Engineering),2014,40(11):3452-3460.
[13]赵熙,杨洪耕(Zhao Xi,Yang Honggeng).基于快速独立分量分析的系统侧谐波阻抗计算方法(Method of calculating system-side harmonic impedance based on FastICA)[J].电力系统自动化(Automation of Electric Power Systems),2015,39(23):139-144,151.
[14]谢庆,张丽君,程述一,等(Xie Qing,Zhang Lijun,Cheng Shuyi,et al).快速独立分量分析算法在局放超声阵列信号去噪中的应用(Application of the FastICA algorithm to PD ultrasonic array signal de-noising)[J].中国电机工程学报(Proceedings of the CSEE),2012,32(18):160-166.
[15]季忠,金涛,杨炯明,等(Ji Zhong,Jin Tao,Yang Jiongming,et al).虚拟噪声通道在基于ICA消噪过程中的应用(Application of virtual noise channels in the processing of signal de-noising with ICA)[J].中国机械工程(China Mechanical Engineering),2005,16(4):350-352,364.
[16]焦卫东(Jiao Weidong).基于独立分量分析的旋转机械故障诊断方法研究(Research on Methods for Fault Diagnosis of Rotating Machines Based on Independent Component Analysis)[D].杭州:浙江大学机械与能源工程学院(Hangzhou:College of Mechanical and Energy Engineering,Zhejiang University),2003.
[17]陈东方,吴先良(Chen Dongfang,Wu Xianliang).采用EMD方法消除瞬态散射回波中的高斯白噪声干扰(Recovery of signal from transient scattered response contaminated by Gaussian white noise based on EMD method)[J].电子学报(Acta Electronica Sinica),2004,32(3):496-498.
[18]钱勇,黄成军,陈陈,等(Qian Yong,Huang Chengjun,Chen Chen,et al).基于经验模态分解的局部放电去噪方法(Denoising of partial discharge based on empirical mode decomposition)[J].电力系统自动化(Automation of Electric Power Systems),2005,29(12):53-56,60.