基于VMD样本熵和CS-ELM的滚动轴承故障诊断

Rolling Bearing Fault Diagnosis Based on VMD Sample Entropy and CS-ELM

为解决滚动轴承故障信号不稳定和故障识别准确率低的问题,结合VMD样本熵特征提取与布谷鸟搜索(CS)算法优化的超限学习机(ELM)进行故障识别实验。首先对故障信号进行VMD分解并计算样本熵形成特征向量,然后通过CS算法优化ELM输入权值和隐含层阈值,最后利用CS-ELM模型进行分类诊断。实验结果表明:基于VMD样本熵和CS-ELM的滚动轴承故障识别准确率高于99%。

For purpose of eradicating instability of the rolling bearing fault signals and improving fault recognition accuracy,the over-limit learning machine(CS-ELM)optimized by the VMD(variational mode decomposition)sample entropy feature extraction and the cuckoo search algorithm were adopted for fault recognition to improve accuracy of the rolling bearing diagnosis.Firstly,having the fault signals decomposed with VMD and the sample entropy calculated to form feature vector;and then,having the ELM weights and thresholds optimized with CS algorithm;and finally,having the CS-ELM model employed for classification and diagnosis.The experimental results show that,the fault recognition accuracy based on VMD sample entropy and CS-ELM model is above 99%.