基于神经网络的异步电动机信息融合故障诊断系统

针对传统异步电动机故障诊断方法中存在的局限性,在对异步电动机故障诊断的特点和要求基础上,提出了一种基于神经网络的信息融合故障诊断方法.对所采集异步电动机的电压、电流、绕组温度等进行数据预处理与特征提取、归一化后,把这些特征参数作为神经网络的输入,经过学习训练,以判断系统状态,识别系统的故障.仿真实验结果表明其故障诊断是可行和有效的.

Fault Diagonosis System for Asynchronous Motor Based on Neural Netork Information Fusion

Aiming at limitation of the traditional method of fault diagnosis for asynchronous motor, based on the research of the characteristics and demands of fault diagnosis for asynchronous motor, this paper puts forward a fault diagnosis method base on neural network information fusion. This method used characteristic information of asynchronous motor such as voltage, current, winding temperature, etc, finishes data preprocessing, feature extraction, and normalization. Then it uses these characteristic pa- rameters as the inputs of the neural network, studies and trains, judges the state of system, and recognizes the fault of system. The simulating experiment results of the system based on the method shows that its fault diagnosis is feasible and effective.