基于VMD⁃ESA和IPOA⁃XGBOOST相结合的异步电机故障诊断

Asynchronous motor fault diagnosis based on combination of VMD⁃ESA and IPOA⁃XGBOOST

为了提高异步电机故障诊断的准确度,提出一种结合变分模态分解(VMD)、包络谱分析法(ESA)和改进的鹈鹕优化算法优化的极限梯度提升模型(IPOA‐XGBOOST)的智能诊断方法。首先,对实测的异步电机振动信号进行VMD分解,并用ESA计算VMD分解得到的本征模态分量(IMFs)的瞬时能量矩阵;然后用奇异值分解法(SVD)对得到的瞬时能量矩阵进行特征提取;最后,使用提取到的特征向量训练IPOA‐XGBOOST模型,得到异步电机的故障诊断准确率。另外,为了解决鹈鹕优化算法容易陷入局部最优解、寻优速度慢等问题,使用Circle映射改进鹈鹕优化算法。将改进的鹈鹕优化算法、遗传算法(GA)和鹈鹕优化算法进行寻优分析,实验结果表明,改进的鹈鹕优化算法的寻优效果最好。

In order to improve the accuracy of fault diagnosis of asynchronous motor,an intelligent diagnostic method is proposed that combines variational mode decomposition(VMD),envelope spectrum analysis(ESA),and improved Pelican optimization algorithm to optimize the limit gradient lifting model(IPOA‐XGBOOST).VMD decomposition on the measured vibration signals of asynchronous motors is conducted,and ESA is used to calculate the instantaneous energy matrix of the intrinsic mode components(IMFs)obtained from VMD decomposition.Then,singular value decomposition(SVD)is used to extract features from the obtained instantaneous energy matrix.The extracted feature vectors are used to train the IPOA‐XGBOOST model and obtain the accuracy of fault diagnosis of asynchronous motor.In order to solve the problems that the pelican optimization algorithm is easy to fall into the local optimal solution and the optimization speed is slow,the Circle mapping is used to improve the pelican optimization algorithm.The optimization analysis of improved pelican optimization algorithm,genetic algorithm(GA)and pelican optimization algorithm are conducted.The experimental results show that the improved pelican optimization algorithm has the best optimization effect.