【摘要】 针对低频振动信号 ,给出一种故障信息识别的有效方法 .选择双正交样条小波作为基小波 ,将多尺度分析及孤立奇异点的检测运用到水轮机主轴径向摆度分量信号的故障识别中 .结果表明 ,利用连续小波变换系数模极大值 (WTMM)的多尺度分析、根据WTMM曲线的长度、强度及其Lipschitz指数可以定位孤立奇异点 ;运用最小二乘算法估算李氏指数 ,即可估算故障点的奇异性程度并取得了很好的诊断效果更多还原

【Abstract】 As for low frequency vibrating signal, an effective method of identifying fault information is put forward. With biorthogonal spline wavelet selected, fault points of the main shaft of hydroelectric generator can be recognized, using MRA(multi resolution analysis ) theory of CWT (continuous wavelet transform) and isolated singularity points detecting. A perfect method is designed that positions the isolated singularity points and estimates theirs singularity degree, basing on WTMM (wavelet transform maximal modulus), MRA of wavelet transform and Least Square Algorithm, taking account of the length, intensity, and Lipschitz exponent of WTMM curves.更多还原