基于小波能量熵和粗糙集的振动故障诊断研究

摘要

In this paper,Analyze the vibration fault signal with wavelet method to get a number of wavelet coefficients,and then deducting redundant characteristics which get from the wavelet energy based on rough set, To diagnose the fault samples by the generated diagnostic rules form the reduction characteristics, On the basis of the fault have been identified, Measure the fault degree with wavelet energy spectrum entropy. The experiment results show that the RSES software based on rough set theory have a higher diagnostic accuracy,To a certain extent the fault degree can be measured through energy spectral entropy parameters.%通过对不同振动故障信号进行小波包分解,得到若干个小波分解系数,进而从中获得各小波系数能量,并以此为特征进行基于粗糙集理论的特征约简分析,根据约简后生成的诊断规则对故障样本进行诊断,在确定故障的基础上,通过小波能量谱熵来衡量故障严重程度.实验结果显示通过基于粗糙集理论的RSES软件实现故障诊断准确率较高,同时能量谱熵参数可以在一定程度上衡量故障程度.