基于贝叶斯优化的极限梯度上升树方法对电机轴承故障诊断方法的研究

Research on Bearing Fault Diagnosis Method of XGboost Motor based on Bayesian Optimization

针对电机轴承故障问题,提出一种基于极限梯度上升树(XGboost)与贝叶斯优化(Bayesian Optimization)相结合的电机轴承故障诊断方法。提取电机轴承振动信号的小波包能量特征,使振动信号具有较好的可靠性,提高了故障诊断的准确率。采用贝叶斯优化算法对极限梯度上升树(XGboost)中的最大迭代次数、上升树的最大深度等参数进行超参数优化,并与故障诊断中常用的其他算法进行对比。实验结果表明:基于贝叶斯优化的极限梯度上升树(XGboost)的方法不仅能够实现对电机轴承的不同位置故障的准确识别,而且对每一个位置故障的严重程度有较好的诊断效果,具有较强的实用性。

Aiming at the problem of motor bearing failure, a method of motor bearing fault diagnosis based on the combination of XGboost and Bayesian optimization is proposed. Extracting the wavelet packet energy characteristics of the vibration signal of the motor bearing makes the vibration signal have better reliability and improves the accuracy of fault diagnosis. The Bayesian optimization algorithm is used to optimize the parameters such as the maximum number of iterations and the maximum depth of the tree in XGboost, and compare with other algorithms commonly used in fault diagnosis. The experimental results show that the Bayesian-optimized limit gradient ascending tree(XGboost) method can not only realize the accurate identification of the faults at different positions of the motor bearings, but also have a good diagnosis effect for the severity of each position fault, and has astrong practicality.