首页 | 本学科首页   官方微博 | 高级检索  
     

基于KPCA与LSSVM的惯性测量组合模拟电路故障诊断
引用本文:冯磊,王宏力,侯青剑. 基于KPCA与LSSVM的惯性测量组合模拟电路故障诊断[J]. 战术导弹技术, 2009, 0(4): 76-80
作者姓名:冯磊  王宏力  侯青剑
作者单位:第二炮兵工程学院,西安710025
摘    要:针对惯性测量组合模拟电路因容差造成的软故障不易诊断的问题,将核主元分析与最小二乘支持向量机结合,应用于惯性测量组合模拟电路软故障的诊断.通过对电路频率响应的输出波形进行分析,选取不同频率下的电压值作为原始特征,利用核主元分析提取主要特征,然后利用最小二乘支持向量机对各种状态下的特征向量进行分类决策,实现惯性测量组合模拟电路的软故障诊断.仿真结果表明,该方法计算简单,能够准确诊断软故障,具有较高的识别率.

关 键 词:核主元分析  最小二乘支持向量机  惯性测量组合  模拟电路  故障诊断

Fault Diagnosis of Analog Circuit in Inertial Measurement Unit Based on KPCA and LSSVM
Feng Lei,Wang Hongli,Hou Qingjian. Fault Diagnosis of Analog Circuit in Inertial Measurement Unit Based on KPCA and LSSVM[J]. Tactical Missile Technology, 2009, 0(4): 76-80
Authors:Feng Lei  Wang Hongli  Hou Qingjian
Affiliation:(The Second Artillery Engineering College, Xi'an 710025, China)
Abstract:A method of fault diagnosis based on KPCA and LSSVM is presented and is used for the fault diagnosis of analog circuit in inertial measurement unit ( IMU ). The output waveform of frequency response in the circuit is analyzed, and the voltage value under different frequency is used as original feature. The input data is preprocessed by KPCA, and the feature vectors are classified by LSSVM under certain states. The fault diagnosis in analog circuit is realized. The simulation results show that the algorithm is simple, and it can diagnose the fault of analog circuit correctly.
Keywords:kernel principal component analysis  least squares support vector machine  analog circuit  fault diagnosis
本文献已被 维普 等数据库收录!
设为首页 | 免责声明 | 关于勤云 | 加入收藏

Copyright©北京勤云科技发展有限公司  京ICP备09084417号