基于最优小波基的主泵裂纹转子特征识别研究

Feature recognition of cracked rotor of RCP based on optimal wavelet basis

针对反应堆主冷却剂泵经过长期运行后,可能会出现的转子裂纹故障,利用小波分析,能够实现对故障特征的识别。结合反应堆主冷却剂泵裂纹转子振动模型的仿真信号,运用连续小波变换的方法,从小波基库中选出不同小波基分别计算与故障信号的互相关系数,确定最大值,其对应的小波基即为转子裂纹故障信号的最优小波基,并使用该小波基对故障信号进行分析,将功率谱、变换尺度以及频率以三维图的形式刻画出来。仿真结果表明,最大互相关系数选出的小波基可以作为转子裂纹故障信号的最优小波基,并且能够很好的识别出转子裂纹故障特征。

The crack fault of its rotor may occur after a RCP operates for a long time,the fault feature can be identified effectively with the method of wavelet analysis.Based on the simulation signals of a vibration model of a RCP’s cracked rotor and the method of CWT,the cross-correlation coefficients between the fault signals and the different wavelet bases selected from a wavelet basis library were computed here.After determining the maximum value of the cross-correlation coefficients,the corresponding wavelet basis was the optimal wavelet basis for the fault signals of the cracked rotor.The optimal wavelet basis was used to analyze the fault signals.A three-dimensional diagram with respect to power spectrum,scale of transformation and frequency was presented.The simulation results illustrated that the wavelet basis selected with the maximum cross-correlation coefficient can be taken as the optimal wavelet basis,and with it the fault feature of the cracked rotor can be recognized effectively.