Abstract:GIS partial discharge pattern recognition is an important part of its state evaluation, and a 252kV GIS partial discharge detection simulation experiment platform based on UHF method, and four kinds of typical partial discharge models in laboratory are designed, then corresponding UHF signal mapping database through the experimental method, and also extracted the original 26 feature parameters; the authors reduced the dimension of the feature space through principal component analysis, and then obtain 10 new characteristics parameters, inputed the original features and the characteristic parameters after dimension reduction to the classification of relevance vector machine classifier respectively. The results show that taking the characteristic parameters after the dimension reduction as input, the recognition rate is higher than the original feature parameters; BN, SVM adopt to do the comparison with milticlass relevance vector machine, The results show that input the original characteristic parameters or the dimension reduction parameters, the M-RVM method recognition rate is highest, and the recognition rate is above 85% after the dimension reduction.
[1] 司文荣, 李军浩, 袁鹏, 等. 基于波形非线性映射的多局部放电脉冲群快速分类[J]. 电工技术学报, 2009, 24(3): 217-228. Si Wenrong, Li Junhao, Yuan Peng, et al. The fast grouping technique of PD sequence based on the nonlinear mapping of pulse shapes[J]. Transactions of China Electrotechnical Society, 2009, 24(3): 217- 228. [2] 李信, 李成榕, 丁立健, 等. 基于特高频信号检测GIS局放模式识别[J]. 高电压技术, 2003, 14(3): 16-20. Li Xin, Li Chengrong, Ding Lijian, et al. Identifica- tion of PD patterns in gas insulated swichgear(GIS) based on UHF signals[J]. High Voltage Engineering, 2003, 14(3): 16-20. [3] 唐炬, 周倩, 许中荣, 等. GIS特高频局放信号的数学建模[J]. 中国电机工程学报, 2005, 25(19): 106- 110. Tang Ju, Zhou Qian, Xu Zhongrong, et al. Establish- ment of mathematical model for partial discharge in GIS using UHF method[J]. Proceedings of the CSEE, 2005, 25(19): 106-110. [4] Judd M D, Cleary G P, Bennoch C J. Applying UHF partialdischarge detection to power transformers[J]. IEEE Power Engineering Review, 2002, 22(8): 57-59. [5] 司文荣, 李军浩, 袁鹏, 等. 气体绝缘组合电器多局部放电源的检测与识别[J]. 中国电机工程学报, 2009, 29(16): 119-125. Si Wenrong, Li Junhao, Yuan Peng, et al. Detection and identification techniques for multi-PD sources in GIS[J]. Proceedings of the CSEE, 2009, 29(16): 119-125. [6] 胡文堂, 高胜友, 余绍峰, 等. 统计参数在变压器局部放电模式识别中的应用[J]. 高电压技术, 2009, 35(2): 277-281. Hu Wentang, Gao Shengyou, Yu Shaofeng, et al. Application of statistic parameters in recognition of partial discharge in transformers[J]. High Voltage Engineering, 2009, 35(2): 277-281. [7] Gulski E, Kreuger F H. Computer-aided recognition of discharge sources[J]. IEEE Transactions on Electrical Insulation, 1992, 27(1): 82-92. [8] Mazroua A A. Discrimination between PD pulse shapes using different neural network paradigms[J]. IEEE Transactions on Dielectrics and Electrical Insulation, 1994, 1(6): 1119-1131. [9] 张晓虹, 张亮, 乐波, 等. 基于局部放电的矩特征分析大电机主绝缘的老化[J]. 中国电机工程学报, 2002, 22(5): 94-98. Zhang Xiaohong, Zhang Liang, Le Bo, et al. Analysis onaging condition of stator winding insulation of generator basedon the moment characteristics of partial discharge[J]. Proceedings of the CSEE, 2002, 22(5): 94-98. [10] 廖瑞金, 杨丽君, 孙才新, 等. 基于局部放电主成分因子向量的油纸绝缘老化状态统计分析[J]. 中国电机工程学报, 2006, 26(14): 114-119. Liao Ruijin, Yang Lijun, Sun Caixin, et al. Aging conditionassessment of oil-paper based on principal component and fac-tor analysis of partial discharge[J]. Proceedings of the CSEE, 2006, 26(14): 114-119. [11] Damoulas T, Girolami M A. Probabilistic multi-class multi-kernel learning: on protein fold recognition and remote homology detection[J]. Bioinformatics, 2008, 24(10): 1264-1270. [12] Girolami M, Rogers S. Hierarchic Bayesian models for kernel learning[C]. In Proceedings of the 22nd International Conference on Machine Learning, Bonn, Germany, 2005, 241-248. [13] 李信, 李成榕, 丁立健, 等. 基于特高频信号检测GIS局放模式识别[J]. 高电压技术, 2003, 14(3): 16-20. Li Xin, Li Chengrong, Ding Lijian, et al. Identifica- tion of PD patterns in gas insulated swichgear(GIS) based on UHF signals[J]. High Voltage Engineering, 2003, 14(3): 16-20. [14] Psorakis I, Damoulas T, Girolami M A. Multiclass relevance vector machines: sparsity and accuracy[J]. IEEE Transactions on Neural Networks, 2010, 21(10): 1588-1598. [15] Damoulas T, Girolami M A. Combining feature spaces for classification[J]. Pattern Recognition, 2009, 42(11): 2671-2683.