一种基于参数优化VMD联合融合特征的电机轴承故障诊断方法

打开文本图片集

AMethod for Motor Bearing Fault DiagnosisBased on ParameterOptimized VMD and Joint Fusion Features

JIN Guoqing CAO Zhenguan （Anhui UniversityofScienceand Technology，Huainan232Ooo，China）

Abstract：[Purposes] To address the difficulties in feature extraction and low diagnostic accuracy for motor bearing faults，an inteligent diagnosis method based on parameter-optimized Variational Mode Decomposition （VMD） and fusion features is proposed to improve fault characterization and recognition accuracy.[Methods] ① Parameter optimization： The Sparrow Search Algorithm （SSA） is introduced to adaptively optimize VMD's modal number K and penalty factor α ，solving the problem of manual selection （SSA-VMD）. ② Feature extraction： Bearing vibration signals are decomposed using SSA-VMD to generate Intrinsic Mode Function （IMF） sets; the optimal IMF components are selected based on minimized local envelope entropy，and their energy proportions and energy entropy are calculated to construct a feature vector matrix. ③ Classification diagnosis：The feature matrix is input into Support Vector Machine （SVM），and the Golden Jackal Optimization （GJO） algorithm is used to adaptively optimize SVM's penalty factor and kernel function parameters to establish the GJO-SVM diagnostic model.[Findings] Experiments show that the proposed method （SSA-VMD fusion features + GJO-SVM） achieves a fault diagnosis accuracy rate of 98.75% ，significantly outperforming comparison methods such as Particle Swarm Optimization-based SVM （PSO-SVM） and Fruit Fly Optimization Algorithm-based SVM （FOA-SVM）. [Conclusions] This method realizes automatic parameter optimization of VMD through SSA，enhances fault characterization using energy and entropy fusion features，and improves SVM classification performance with GJO，forming a high-precision diagnostic system that provides a reliable technical solution for industrial bearing condition monitoring.

Keywords： parameter optimization VMD; fusion feature; golden jackal optimization; support vector machine; bearing fault diagnosis

0 引言

在电机运行状态的可靠性评估中，轴承工况是关键指标之一。（剩余5857字）