变模式分解和熵值法的轴承故障诊断

Bearing Fault Diagnosis Based on Variable Mode Decomposition and Entropy Value Method

针对峭度等单一指标筛选最优频带所存在的困难,提出了利用熵值法计算多指标的权重,将多个与滚动轴承故障有关的特征指标进行融合。对轴承振动信号进行变模式分解,再对多个频带进行筛选,后得出最优频带,并对最优频带进行频谱分析。通过实际信号对该方法进行了验证,结果表明该方法能准确判明轴承运行状态。

Aiming to the difficulty in finding the optimum frequency band using single index, for example kurtosis, it is proposed to apply multi index weight calculated by entropy the method, and then fuse these multiple characteristic index for rolling bearing fault detection. The vibration signal is decomposed by variable mode decomposition, and then the optimal frequency band is selected. The frequency analysis is conducted to the optimal frequency band. The results show that the bearing operation condition can be recognized accurately by the proposed method.