【摘要】 研究了离散小波变换的若干特性 ,以及用信息熵来反映信号统计分布特征的方法 ,引进了多尺度下小波熵和相对小波熵的概念 ,并且以此为参量对系统状态进行趋势分析 ,从而突出系统信号中短暂的异常信号 ,达到及早发现可能故障的目的。最后通过数值实验和应用实例说明 :分析这两个参数的变化情况 ,可以反映系统的状态趋势 ,提高维修的针对性 ,降低设备维修的代价 ,提高设备的安全性和可靠性更多还原

【Abstract】 In this article some characteristics of discrete wavelet transform are analyzed, and the entropy-based method is developed to reflect the statistic distribution of signals. Furthermore, the concept of wavelet entropy and relative wavelet entropy at different scales are introduced, which are used to analyze the trends of mechanical systems. Using this method, the ephemeral abnormal information in system signals can be detected in incipient faults. At last, a numeric and case study demonstrates that the proposed method can reflect the trends of mechanical systems, help diagnosis the fault in repair, cut down the maintanace cost of equipment and improve the safety and reliability of the equipment.更多还原