5 结论
(1)由于超声波信号的检测既不受电磁干扰,又可以避免气相色谱法的滞后问题,利用超声波信号进行在线的局部放电模式识别是可行的。
(2)利用分形理论提取局部放电超声波信号的分形参数可以很好地对其进行局部放电的特征提取。
(3)可以利用人工神经网络对所提取的特征进行模式识别。
(4)对变压器局部放电所产生的超声波信号进行局部放电模式识别。在理论上为变压器局部放电的模式识别提供了一种新的方法。
参考文献
[1] Krivda A. Automated recognition of partial discharge [J]. IEEE Transactions on Dielectrics and Electrical Insulation, 1995, 2(5):796-812.
[2] 董旭柱,王昌长,朱德恒(Dong Xuzhu,Wang Changchang, Zhu Deheng). 电力变压器局部放电在线监测研究的现状与趋势(一)(The status and trend of on-line detection of partial discharge in power transformers) [J].变压器(Transformers),1996,33(1):3-7.
[3] 谢和平,薛秀谦(Xie Heping,Xue Xiuqian ). 分形应用中的数学基础与方法(The mathematical base and method of fractal application) [M]. 北京:科学出版社(Beijing:Science Press),1998.
[4] Satish L, Zaengl W S. Can fractal features be used for recognizing 3-d partial discharge patterns? [J]. IEEE Transactions on Dielectrics and Electrical Insulation, 1995, 2(3):352-359.
[5] Krivda A, Gulski E. The use of fractal features for recognition of 3-D discharge patterns[J]. IEEE Transactions on Dielectrics and Electrical Insulation, 1995, 2(5):889-892.
[6] 姜磊,朱德恒,李福祺,等(Jiang Lei,Zhu Deheng,Li Fuqi ,et al). 基于人工神经网络的变压器绝缘模型放电模式识别的研究(Discharge pattern recognition of insulation model of electrical transformers based on ANN) [J]. 中国电机工程学报(Proceedings of the CSEE),2001,21(1):21-24.
[7] 淡文刚,陈祥训,郑健超(Dan Wengang,Chen Xiangxun,Zheng Jianchao). 油中局部放电脉冲波形的测量与特性分析(Measurement and analysis of pulse current of partial discharge in oil)[J]. 电网技术(Power System Technology),2000,24(6):37-40.
[8] 王泽忠,李成榕,彭湃,等(Wang Zezhong,Li Chengrong,Peng Pai,et al). 基于分形方法的电力设备绝缘局部放电识别(The recognition of PD of electric apparatus insulation based on fractional methods) [J]. 现代电力(Modern Electric Power) ,1999,16(4):33-37.
[9] 丁庆海,庄志洪,祝龙石,等(Ding Qinghai,Zhuang Zhihong,Zhu Longshi,et al). 混沌、分形和小波理论在被动声信号特征提取的应用(Application of the chaos, fractal and wavelet theories to the feature extraction of passive acoustic singal)[J]. 声学学报(Acta Acustica),1999,24(2):197-203.