基于小波域隐马尔可夫模型故障诊断方法  

Bearing fault diagnosis using a wavelet-domain hidden markov model

作　　者：陶新民[1] 徐晶[2] 杜宝祥[1] 徐勇[1] 

机构地区：[1]哈尔滨工程大学信息与通信工程学院,哈尔滨150001 [2]黑龙江科技学院数力系,哈尔滨150027

出　　处：《振动与冲击》2009年第4期33-37,201,共5页Journal of Vibration and Shock

基　　金：哈尔滨工程大学校科研基金(0020802600735);黑龙江省博士后基金(3236301199)

摘　　要：针对基于小波能量谱和能量谱熵的故障诊断方法要求小波分解系数基本符合高斯分布这一不足,提出一种基于多尺度小波域隐马尔可夫模型(WHMM)参数特征的故障诊断方法。该方法分析了信号多尺度小波分解系数的统计特征,利用隐马尔可夫模型描述小波变换域系数在尺度间,尺度内的统计相关性。采用最大似然估计方法确定的模型参数作为信号特征实现故障诊断。试验结果证实了设计思想的正确性和算法的高效检测性能。最后从小波基、窗口宽度和分类器三个层面对建议方法诊断性能的影响进行分析,结果表明本文方法具有很强的稳定性和鲁棒性。In order to avoid insufficient characteristics of traditional wavelet energy spectrum for bearing fault detection, a novel fault detection method based on wavelet histogram signatures which could capture all the first order statistics using a model based on hidden Markov model (HMM) was presented. In this approach, the statistical features of the multi-scale wavelet coefficients generated by wavelet decomposition of a signal were analyzed, the detail wavelet histogram of a bearing vibration signal could be modeled using the hidden Markov model. The parameters of this model as diagnosis features were introduced to completely describe the wavelet coefficients’ first-order statistics. The scale and shape parameters of the model were estimated with the maximum likelihood method. Comparison of the performance of the proposed approach with those of the methods based on the wavelet energy spectrum and the wavelet energy spectrum entropy was done. The results showed the relative effectiveness of the introduced feature sets in a bearing condition detection with some concluded remarks. The effects of wavelet base selection, window width and classifier on the proposed method were studied in experiments, they could evaluate the stability and robustness of the proposed method.

关 键 词：故障诊断 隐性马尔可夫模型 小波能量谱 最大似然估计方法 

分 类 号：TM711[电气工程—电力系统及自动化]