以盒维数和信息维数为识别特征量的GIS局部放电模式识别方法  

PD PATTERN RECOGNITION METHOD USING BOX DIMENSION AND INFORMATION DIMENSION AS DISCRIMINATING FEATURES IN GIS

作　　者：孙才新[1] 许高峰[1] 唐炬[1] 陆宠惠[1] 侍海军[1] 

机构地区：[1]重庆大学高电压与电工新技术教育部重点实验室,重庆市沙坪坝区400044

出　　处：《中国电机工程学报》2005年第3期100-104,共5页Proceedings of the CSEE

基　　金：国家自然科学基金项目(50377045)~~

摘　　要：研制了5种典型的GIS人工模拟缺陷模型及其局部放电检测系统,通过实验获取了大量局部放电样本数据,构造出GIS局部放电灰度图象;提出有效估计图象盒维数的最少盒计数法;提取了GIS局部放电灰度图象的分形特征——盒维数和信息维数;利用以局部放电灰度图象的盒维数和信息维数作为识别特征量,径向基函数人工神经网络为识别分类器的GIS局部放电模式识别方法,取得了良好的识别效果。Five kinds of typical artificial defect models in Gas Insulated Substation(GIS) and PD detection system were developed. PD sample data caused by typical artificial defects in GIS were got by PD detection system in laboratory. Utilizing these sample data, GIS PD gray intensity images were constructed. Minimum box-counting method which is used to estimate image fractal dimension was presented. Fractal features of GIS PD gray intensity images such as box dimension and information dimension are extracted. GIS PD pattern recognition scheme was put forward. In this scheme, PD fractal dimensions were used as inputs, Radial Basis Function(RBF) neural network was used as classifier. Good recognition result is obtained using the scheme.

关 键 词：局部放电 灰度图象 GIS 计数法 径向基函数 人工神经网络 人工模拟 模式识别方法 信息维数 分类器 

分 类 号：TM41[电气工程—电器] TP301.6[自动化与计算机技术—计算机系统结构]