基于CEEMD和WOA_LSSVM滚动轴承声信号故障诊断

针对强背景噪声下滚动轴承故障诊断问题,结合互补集合经验模态分解(CEEMD)与鲸鱼优化算法优化最小二乘支持向量机(WOA_LSSVM)进行滚动轴承的故障诊断研究。首先对声信号进行快速谱峭度分析并进行带通滤波预处理,提取故障冲击成分;其次,利用CEEMD算法将滤波信号进行分解运算,得到一系列模态分量(IMF);再利用相关系数法选取有效IMF分量进行信号重构;再提取重构信号的近似熵、峭度、峰峰值、峰值因子、波形因子作为特征值组成特征向量;最后,将归一化的特征向量输入WOA_LSSVM进行故障类别识别。将该方法用于滚动轴承试验数据,并进行对比试验分析,验证了该方法的有效性,提高了故障诊断的准确率。

Fault Diagnosis of Rolling Bearing Acoustic Signal Based on CEEMD and WOA_LSSVM

Aiming at the problem of rolling bearing fault diagnosis under strong background noise,this paper combines the complementary set empirical mode decomposition(CEEMD)and the whale optimization algorithm optimization least squares support vector machine(WOA_LSSVM)to research the rolling bearing fault diagnosis.First,perform fast spectral kurtosis analysis of the acoustic signal and perform band-pass filtering pre-processing to extract fault shock components;Then use CEEMD algorithm to decompose the filtered signal to obtain a series of intrinsic mode functions(IMF);then use the correlation coefficient method to select effective IMF components for signal reconstruction;Extract the approximate entropy,kurtosis,peak-to-peak value,crest factor,and waveform factor of the reconstructed signal as the feature values to form a feature vector;Finally,the normalized feature vector is input to WOA_LSSVM for fault category identification.The method was applied to the rolling bearing test data,and a comparative test analysis was performed to verify the effectiveness of the method and improve the accuracy of fault diagnosis.