Study of Health Monitoring System of Machines by Using Wavelet Transform

从保证设备安全和制定设备维修方案的角度出发，估算设备在各种动载荷作用下的损坏率是十分重要的。本研究主要探讨以健康监测系统来监测机器的损坏，以确保机器的安全性。将小波变换应用于机械系统的振动响应分析中，验证了其可行性.从这项研究中获得的结果总结如下。(1)本文用边界元法同时分析了波在由气体场和固体场组成的多层介质中的传播。通过分析，利用响应信号的小波变换，识别出机器异常信号的位置。(2)Daubechies小波的大N值正交小波能在噪声条件下对异常信号的出现时间给出较好的估计。(3)在用小波分析估计进入塑性区的数量时，Meyer小波是三种分析小波中最合理的。当系统的滞回恢复力变化较大时，进入塑性区的数量可以合理估计。

It is very important to estimate the damage rate of machines under the verious dynamic loading from the point of view of guranteeing their safety and developing the scheme of their maintenance. In this study, the health monitoring system to monitor the damage of machines is investigated to assure the safety of machines. The wavelet transform is applied to the analyzed responses of machines and its feasibility is confirmed. The results obtained from this research are summarizeda as follows. (1) Simultaneous analysis of waves propagating through multilayred media consisting of gas and solid fields is done by using the boundary element method. Through this analysis, the location of abnormal signal for machines is identified by using the wavelet transform of responses. (2) The orthonormal wavelet with large N for Daubechies wavelet can give the good estimation for the occurrence time of abnormal signal under the noisy condition. (3) In case of estimating number entering into plastic region by wavelet analysis, Meyer's wavelets are most reasonable among three types of analyzing wavelets. Moreover, as the change of hysteretic restoring force of system becomes large, number entering into plastic region can be estimated remarkablely.