Analog Circuit Soft Fault Based on Fault-test Dependency Matrix and SVM
Recently, with the rapid development of electronic industry, the soft fault problems in analog circuit have caused wide public concern.In order to enhance effectiveness of fault features extraction and validity of soft fault diagnosis in analog circuit, a fault diagnostic method on the basis of fault-test dependency matrix and support vector machine is proposed in this paper.In this method, through sensitivity analysis on output characteristic of analog circuit, the components which are more likely to lead to soft fault are determined;the fault value of the fault components is determined through the failure threshold of output characteristics;fault-test dependency matrix is set up to determine fault characteristics requiring extraction;considering parameter fluctuations of components, Monte Carlo analyses is used to get the set of fault signals;automatic diagnosis of soft fault is realized based on support vector machine.The method is applied to a certain electronic module in this paper to testify its efficiency.
aerospace analog circuit fault diagnosis fault-test dependency matrix support vector machine
Bokai Zheng Jie Deng Kaixin Zhang Qisen Sun Lu Wang Guofu Zhai
Department of Electrical Engineering and Automation, Harbin Institute of Technology Harbin, Heilongjiang,China, 150001
国际会议
江苏常熟
英文
283-289
2017-04-18（万方平台首次上网日期，不代表论文的发表时间）