基于多维特征与IGWO-SVM的电机轴承故障诊断

Motor bearing fault diagnosis based on multi-dimensional features and IGWO-SVM

针对电机轴承故障诊断精度低、传统灰狼优化算法(GWO)优化支持向量机(SVM)故障诊断模型容易陷入局部最优的问题,引入非线性收敛因子和Levy飞行策略对改进灰狼优化算法(IGWO)进行研究,提出了一种基于多维特征与改进灰狼优化算法优化支持向量机(IGWO-SVM)的电机轴承故障诊断方法。提取电机轴承振动信号的时域和频域特征构成多维特征矩阵;采用主成分分析(PCA)降低特征矩阵的数据维数,以实现快速数据处理;利用IGWO对SVM模型参数进行优化,得到最优的IGWO-SVM故障诊断模型用于确定电机轴承的故障类型。实验结果表明:所提出的电机轴承故障诊断方法在不同工况下精度高、性能稳定,所提出的IGWO算法与传统GWO和基于差分进化的改进灰狼优化算法(DEGWO)相比,具有更好的收敛性和精度。

Aiming at the problems of low fault diagnosis accuracy of motor bearings,the fault diagnosis model optimized by traditional grey wolf optimization algorithm(GWO)to support vector machine(SVM)is prone to fall into local optimality.Nonlinear convergence factor and Levy flight strategy are introduced to study the improved grey wolf optimization algorithm,and a motor bearing fault diagnosis method based on multi-dimensional features and improved grey wolf optimization algorithm optimized support vector machine(IGWO-SVM)is proposed.The time domain and frequency domain characteristics of motor bearing vibration signals are extracted to form a multidimensional characteristic matrix.The principal component analysis(PCA)is used to reduce the data dimension of feature matrix to realize fast data processing.IGWO is used to optimize the parameters of SVM model,and the optimal IGWO-SVM fault diagnosis model is obtained for determining the fault types of motor bearings.Experimental results show that the proposed motor bearing fault diagnosis method has high accuracy and stable performance under different working conditions.Compared with traditional GWO and improved gray wolf optimization algorithm based on differential evolution(DEGWO),the proposed IGWO algorithm has better convergence and accuracy.