电工技术学报2018，Vol.33Issue(1)：158-165,8.DOI:10.19595/j.cnki.1000-6753.tces.160966
基于小波包优选的模拟电路故障特征提取方法
Fault Feature Extraction Method for Analog Circuit Based on Preferred Wavelet Packet
摘要
Abstract
Wavelet packet transform is widely used in analog circuit fault feature extraction, and the wavelet packet basis affects the performance of the fault feature extraction directly. The paper defined frequency band energy entropy of the wavelet packet, and a wavelet packet selection method was presented based on the frequency band energy entropy and its variance.In order to verify the feasibility of the method,different mother wavelet functionwere used to extract the fault features of the circuit under test,and support vector machinewas presented to verify the features' identifiability. Simulation results showed that the extracted fault features with the selected optimal wavelet had higher discrimination compared with the fault features extractedwith other wavelets, which indicated that the method is feasible.关键词
小波包基函数/频带能量熵/模拟电路/特征提取Key words
Wavelet packet basis/frequency band energy entropy/analog circuit/feature extraction分类
信息技术与安全科学引用本文复制引用
袁莉芬,孙业胜,何怡刚,张悦,吕密..基于小波包优选的模拟电路故障特征提取方法[J].电工技术学报,2018,33(1):158-165,8.基金项目
国家重点研发计划(2016YFF0102200)、中国博士后特别资助项目(2015T80651)、中国博士后面上项目(2014M5517)、合肥工业大学春华计划项目(2014HGCH0012)、国家自然科学基金(61102035, 51577046)和国家自然科学基金重点项目(51637004)资助. （2016YFF0102200）
