一种电机故障的智能诊断方法研究

Research on intelligent diagnosis of motor failure based on IGA-BP mixed algorithm

提出了一种基于改进遗传算法(IGA)和误差反向传播(BP)算法相结合的IGA-BP混合算法在电机故障诊断中的应用,首先提取电机转子振动频谱分量作为神经网络的训练样本,将代表故障信息的数据作为输入量代入已训练好的神经网络后,通过输出结果即可诊断故障类型。仿真实验表明,该方法可以有效地识别电机常见故障,诊断准确率高、速度快。

The IGA-BP mixed algorithm based on the improved genetic algorithm （IGA） and to back propagation （BP） algorithm are presented,which unifies motor failure diagnosis and makes a mixed algorithm to use in training the artificial neural network,carrying on failure diagnonsis for the motor rotor. First withdraws the motor rotor＇s vibrate the frequency spectrum component to take the neural network as the training sample,inputs the failure information data to the neurar network which has trained well already, then diagnoses the failure type through the output result. The simulation results show that the method can disitinguish the common motor failures effectively, and has the characteristics of quick convergence rate and high diagnosis precision.