Abstract：Feature extraction is the key to GIS partial discharge pattern recognition, usually, the dimension of feature space is high, which is not conductive to classification. Based on this, the article introduce the principal component sparse thoughts, first of all, through the 252kV GIS partial discharge simulation experiment platform, set up the typical GIS partial discharge models, and uses ultrasound to obtain the corresponding signals, and then, through the principal component contribution rate to decide the degree of sparse, the results show that using this method can realize effective extraction of characteristic, and enhance the clustering results.
律方成, 金虎, 王子建, 张波. 主分量稀疏化在GIS局部放电特征提取中的应用[J]. 电工技术学报, 2015, 30(8): 282-288.
Lü Fangcheng, Jin Hu, Wang Zijian, Zhang Bo. Principal Component Sparse and Its Application in GIS Partial Discharge Feature Extraction. Transactions of China Electrotechnical Society, 2015, 30(8): 282-288.
[1] 李立学, 黄成军, 曾奕, 等. GIS局部放电包络信号噪声抑制的小波方法[J]. 高压电器, 2009, 45(1): 33-35.Li Lixue, Huang Chengjun, Zeng Yi, et al. Noise reduction using wavelet for envelope signal of partial discharge in GIS[J]. High Voltage Apparatus, 2009, 45(1): 33-35. [2] 律方成，金虎，王子建，等. 基于组合核多特征融合的GIS局部放电检测与识别[J]. 电工技术学报, 2014, 29(10): 334-340.Lü Fangcheng, Jin Hu, Wang Zijian, et al. GIS partial discharge detection and recognition based on the kernel combination and multiple feature fusion method[J]. Transactions of China Electrotechnical Society, 2014, 29(10): 334-340. [3] 张晓星, 孙才新, 唐炬, 等. 基于统计不相关最优鉴别矢量集的GIS局部放电模式识别[J]. 电力系统自动化, 2006, 30(5): 59-62.Zhang Xiaoxing, Sun Caixin, Tang Ju, et al. PD pattern recognition based on optimal sets of statistical uncorre- lated discriminant vectors in GIS[J]. Aotomation of Electric Power Systems, 2006, 30(5): 59-62. [4] 邵伟, 祝丽萍, 刘福国, 等. 对称阵稀疏主成分分析及其在充分降维问题中的应用[J]. 山东大学学报, 2012, 47(4): 116-126.Shao Wei, Zhu liping, Liu fuguo, et al. Sparse principal component analysis for symmetric matrix and applica- tion in sufficient dimension reduction[J]. Journal of Shandong University,2012, 47(4): 116-126. [5] Zou H, Hastie T, Tibshirani R. Sparse principal- component analysis[J]. Journal of Computational and Graphical Statistics, 2006, 15(2): 265-286. [6] Zass R, Shashua A. Nonnegative sparse PCA[J]. Advances in Neural Information Processing Systems, NIPS, 2007(19): 1561-1568. [7] Meng Deyu, Zhao Qian, Xu Zongben. Improve robustness of sparse PCA by L1-norm maximization [J]. Pattern Recognition，2012，45(1)：487-497. [8] 栗茂林, 梁霖, 王孙安. 基于稀疏表示的故障敏感特征提取方法[J]. 机械工程学报, 2013, 49(1): 73-80.Li Maolin, Liang Lin, Wang Sunan. Sensitive feature extraction of machine faults based on sparse representa- tion[J]. Journal of Machanical Engineering, 2013, 49(1): 73-80. [9] 来五星, 轩建平, 史铁林, 等. Wigner-Ville时频分布研究及其在齿轮故障诊断中的应用[J]. 振动工程学报, 2003, 16(2): 247-250.Lai Wuxing, Xuan Jianping, Shi Tielin, et al. Research of wigner-ville time frequency and applicationin detecting gear pinion fault[J]. Journal of Vibration Engineering, 2003, 16(2): 247-250. [10] Cavallini A, Montanari G C, Contin A, et al. A new approach to the diagnosis of solid insulation systems based on PD signal inference[J]. IEEE Electrical Insulation Magazine, 2003, 19(2): 23-30. [11] 王彩雄, 唐志国, 常文治, 等. 一种多源局部放电信号分离方法[J]. 中国电机工程学报, 2013, 33(13): 212-219.Wang Caixiong, Tang Zhiguo, Chang Wenzhi, et al. A method for multi-source partial discharge signals separation[J]. Proceedings of the CSEE, 2013, 33(13): 212-219.
2015, Vol. 30 
