基于小波与分形理论的电力设备局部放电类型识别  

Partial Discharge Classification based on Wavelet and Fractal Theory

作　　者：杜伯学[1] 魏国忠[1] 

机构地区：[1]天津大学电气与自动化工程学院,天津市南开区300072

出　　处：《电网技术》2006年第13期76-80,共5页Power System Technology

摘　　要：根据小波理论建立了表征局部放电脉冲信号的三维时频谱图,该三维谱图综合反映了局放脉冲信号的3个基本特征:时间分量、频率分量和放电能量的分布。采用了分形理论从所建立的三维时频谱图中提取放电特征,并构成识别特征量,采用误差反传神经网络对局部放电信号的类型进行模式识别。试验结果表明,该方法可有效区分局部放电的类型。On the basis of wavelet analysis a threedimensional time-frequency pattern to characterize partial discharge （PD） impulse signal is upbuilt which comprehensively shows three basic features of PD impulse signals： time component, frequency component and distribution of discharging energy. From the upbuilt three-dimensional pattern the discharge features are extracted with fractal theory, thus PD fractal dimensions are used as feature vectors. The pattern recognition of the type of PD signal is conducted by means of BP neural network. The discharge experiments have been conducted to validate the proposed method with five types artificial discharge models, and experiment results show that the proposed method can effectively distinguish the type of PD.

关 键 词：局部放电 特征提取 分形理论 模式识别 高电压绝缘技术 

分 类 号：TM83[电气工程—高电压与绝缘技术]