摘要
为了降低模拟电路参数故障的测试难度,提出了基于自适应小波分解和SVM的模拟电路故障诊断的新方法,该方法对电路故障响应进行小波分解提取最优故障特征,母小波的选择是根据被测电路的正常响应和故障响应小波系数之差的最大均方根原则,并引入支持向量机对故障进行分类识别。小波分解具有自适应性,支持向量机结构简单,泛化能力强。实验结果证明了所提的基于自适应小波分解和SVM的模拟电路故障诊断方法是有效的,其故障诊断率大于96.8%。
In order to reduce difficulty of testing analog circuit parametric fault, a method for diagnosing analog circuit faults using adaptive wavelet analysis and support vector machine （SVM） is presented in this paper. Wavelet analysis is applied to extract the optimal fault feature of the circuit under test （CUT） and the mother wavelet selection criteria is based on the maximum root mean square of the difference between the normal signal wavelet coefficient and the faulty signal wavelet coefficient of the CUT. SVM is introduced to identify the faults of the CUT. Wavelet analysis is adaptive and SVM has the advantages of simple structure and strong generalization ability. Experimental results prove that the proposed method for diagnosing analog circuit faults using wavelet analysis and SVM is effective and the fault diagnosis accuracy of the method is batter than 96.8%.
出处
《仪器仪表学报》
EI
CAS
CSCD
北大核心
2008年第10期2105-2109,共5页
Chinese Journal of Scientific Instrument
基金
国家自然科学基金(60372001
90407007)资助项目
关键词
模拟电路
故障诊断
自适应小波分解
支持向量机
analog circuit
fault diagnosis
adaptive wavelet analysis
support vector machine
