论文部分内容阅读
将小波包多分辨与信息熵相结合,提出了一种故障检测与诊断的方法——小波包特征熵一故障法。首先对采集到的振动信号进行3层小波包分解,在通频范围内得到分布在不同频段内的分解序列,进而建立信号的小波包特征熵向量,选取最能反映故障特征的参数作为特征参数,进行故障诊断识别。以水轮机尾水管压力脉动信号为例,运用此法进行了尾水管动态特性信息提取。试验表明小波包特征熵法是提取故障信息并进行故障识别的一种行之有效的方法,为流体机械的故障诊断开拓了新的思路。
Combining wavelet packet multiresolution and information entropy, a method of fault detection and diagnosis is proposed - wavelet packet feature entropy - fault method. Firstly, the vibration signals collected are decomposed by 3-layer wavelet packet, and the decomposition sequences distributed in different frequency bands are obtained in the pass frequency range, and then the wavelet packet feature entropy vector of the signal is established. The parameters that can best reflect the fault features are selected as the feature parameters , For fault diagnosis identification. Taking the pressure fluctuation signal of the draft tube of hydraulic turbine as an example, this method was used to extract the dynamic characteristic information of the draft tube. Experiments show that the wavelet packet feature entropy method is an effective method to extract fault information and identify faults, which opens up new ideas for the fault diagnosis of fluid machinery.