萤火虫-粒子群优化神经网络的异步电机转子断条故障诊断

Fault Diagnosis for Asynchronous Motor Rotor Broken Bar Based on Glowworm Particle Swarm Optimization and Neural Network

针对目前异步电机转子断条故障诊断方法存在的局限性及其缺陷,在利用小波包分析提取电机转子断条故障特征向量的基础上,提出一种基于萤火虫-粒子群神经网络的故障诊断方法,构建电机转子断条的神经网络故障诊断模型,采用萤火虫-粒子群算法优化神经网络的结构参数。试验分析表明,该方法用于电机转子断条故障诊断,诊断速度快、准确性高、可靠性好。

In view of the limitation and defects of the fault diagnosis method for rotor broken bar of asynchronous motor, based on the wavelet packet analysis and extraction of fault feature vector, a fault diagnosis method based on glowworm particle swarm optimization and neural network was presented, a model of neural network fault diagnosis for motor rotor broken bar was constructed, the structure parameters of the neural network were optimized with glowworm particle swarm optimization algorithm. Experiment analysis showed that, this method was applied to the fault diagnosis for motor rotor broken bar, the diagnosis speed was quick, the accuracy was high, the reliability was good.