基于高阶累积量故障特征提取的化工设备电子电路故障诊断  

Fault diagnosis of electronic circuit based on higher ordercumulants fault feature extraction

作　　者：米鑫 戴国强 王黎 张志强 肖伟 MI Xin;DAI Guoqiang;WANG Li;ZHANG Zhiqiang;XIAO Wei(Beijing Metro Technology Development Co.,Ltd.,Beijing 100160,China)

机构地区：[1]北京地铁科技发展有限公司,北京100160

出　　处：《粘接》2022年第8期146-150,共5页Adhesion

摘　　要：研究设计出化工设备电子电路故障诊断系统,通过分析采样得到化工设备故障电路的特征信号,对化工设备电路的故障进行诊断并定位。化工设备电子电路故障诊断的信号采集模块配置8路采集通道进行采样,并以闭环控制的DC-DC变换电路建立仿真模型。经过高阶累积量(HOC)算法处理得到含有原始数据大部分信息的主成分,并组合成新的化工设备数据信息特征向量,降低了数据维度。实验结果显示,故障诊断技术的硬故障诊断率为100%;对化工设备电路中发生软故障的诊断率最高为100%,最低为93%。The study designed the electronic circuit fault diagnosis system,through analyzing and sampling to get the characteristic signal of the fault circuit,to diagnose and locate the fault of the circuit.The signal acquisition module of fault diagnosis equipment is equipped with 8 acquisition channels for sampling,and the simulation model is established by closed-loop control DC-DC transform circuit.After Higher Order Cumulants(HOC)algorithm processing,the principal components containing most information of the original data are obtained and combined into new feature vectors to reduce the data dimension.The experimental results show that the hard fault diagnosis rate of the fault diagnosis technology in this study is 100%,and the soft fault diagnosis rate in the circuit is the highest 100%,the lowest 93%.

关 键 词：故障诊断 特征信号 信号采集模块 HOC算法 变换电路仿真 

分 类 号：TP39[自动化与计算机技术—计算机应用技术] TQ050.7[自动化与计算机技术—计算机科学与技术]