基于优化随机森林算法的高压断路器故障诊断  

Fault diagnosis of high-voltage circuit breaker based on optimized random forest algorithm

作　　者：宋玉琴[1] 王冰 李超[1] 赵洋[1] Song Yuqin;Wang Bing;Li Chao;Zhao Yang(Electronics and Information Engineering College,Xi'an Polytechnic University,Xi'an 710048,China)

机构地区：[1]西安工程大学电子信息学院

出　　处：《电子测量技术》2018年第21期95-98,共4页Electronic Measurement Technology

基　　金：中国纺织工业联合会科技指导性项目(2017070);2018年西安市科技计划(201805030YD8CG14(17))项目资助

摘　　要：针对高压断路器的故障特性,提出基于精确度加权随机森林的故障诊断算法,该方法利用高压断路器的分合闸电流信号数据集,通过Bootstrap采样方法建立决策树,并根据每棵决策树的分类能力设置其权重,建立精确度加权随机森林模型,最后用测试数据集进行检验,完成对高压断路器机械故障的诊断。诊断结果表明,与单棵决策树和基于C4.5的随机森林算法相比,该算法准确率更高,且不会出现过拟合问题,在解决高压断路器故障诊断方面有很好的应用前景。In view of the fault characteristics of high-voltage circuit breaker,a random forest fault diagnosis algorithm based on precision weighted is proposed.This method uses the data set of switching current signal of high voltage circuit breaker to set up decision tree by Bootstrap sampling method,sets its weight according to the classification ability of each decision tree,and builds random forest model based on precision weighted.The weighted random forest model is finally tested by the test data set to realize the fault diagnosis of the high voltage circuit breaker.Compared with the single decision tree and the random forest algorithm based on the C4.5,the diagnosis results show that's algorithm has higher accuracy and no over fitting problems.It has a good application prospect in solving the fault diagnosis of high-voltage circuit breakers.

关 键 词：高压断路器 故障诊断 决策树 C4.5 随机森林 

分 类 号：TM561[电气工程—电器]