基于神经网络专家系统的大型异步电动机故障诊断

针对大型异步电动机发生故障时 ,故障征兆与故障原因之间复杂的对应关系 ,提出了一种利用神经网络专家系统对电动机进行故障诊断的新方法 ,专家系统作为神经网络的外壳 ,充分发挥了专家系统和神经网络各自的特点。文章介绍了系统的构成、功能、算法及实现。该算法利用MATLAB的神经网络工具箱进行了仿真。结果表明 ,该算法可行 ,可大大提高故障诊断的准确性

Fault Diagnosis of Giant Asynchronous Motor Based on Neural Network Expert System

According to the complicated corresponding relationships between fault symptoms and fault reasons when giant asynchronous motor occurring fault, a new fault diagnosis method of motor using neural network expert system was presented. The expert system works as the shell of neural network, making full use of characteristics of expert system and neural network individual. The paper introduced the composition, function, algorithm and its realization of the system. The algorithm also emulated neural network toolbox with MATLAB. The results showed that the algorithm is feasible and the diagnosis accuracy is improved.