基于随机森林的电力电子电路故障诊断技术  

Application of random forests algorithm to fault diagnosis of power electronic circuit

作　　者：鄢仁武[1] 叶轻舟[1] 周理[1] 

机构地区：[1]福建工程学院信息科学与工程学院,福建福州350108

出　　处：《武汉大学学报（工学版）》2013年第6期742-746,共5页Engineering Journal of Wuhan University

基　　金：福建省高校产学合作科技重大项目(编号:2010H6002);福建省自然科学基金项目(编号:2012J05107);福建省教育厅科技项目(编号:JA12230);福建工程学院基金项目(编号:GY-Z10053;GY-Z11070)

摘　　要：电力电子技术广泛应用于各个工程领域中.研究电力电子电路故障诊断方法,可以准确快速隔离故障,提高系统的工作效率.针对电力电子电路发生故障诊断时存在故障模式多的问题,提出了一种基于随机森林算法的电力电子电路诊断方法.文中对随机森林算法的设计细节以及基于随机森林算法的电力电子电路故障诊断步骤进行了详细阐述,并以三相可控整流电路晶闸管故障诊断为例进行验证.诊断结果表明,和现有的诊断方法相比,基于随机森林算法的电力电子电路诊断方法具有较好的诊断正确率和快速性,在解决电力电子电路故障诊断问题上有很好的实用价值和应用前景.Power electronics technology is widely used in various engineering community. So the research on fault diagnosis method of power electronics circuit is valuable. Based on random forests, a new on-line fault diagnosis method for power electronic circuit is put forward. Although there are several existed tech- niques for faults diagnosis of power electronic circuit, the research on random forests algorithm is meaning- ful and necessary because of its fast execution speed, the characteristic of tree classifier, and high process speed. Design details of random forest classifier and evaluating its performance is investigated in this pa- per. A three-phase silicon controlled rectifier circuit with inductive load is used as an example to illustrate the new approach of fault diagnosis. The results of the application indicate that the method can diagnose faults of three-phase rectifier effectively and obtain good diagnosis performance.

关 键 词：故障诊断 电力电子电路 随机森林 

分 类 号：TM773[电气工程—电力系统及自动化]