基于遗传算法和神经网络的远程减速机故障诊断系统的设计

通过对减速机故障形式的分析,采用BP神经网络建立故障诊断模型,利用遗传算法优化神经网络权值、阈值、网络结构,将遗传神经网络模型应用到远程减速机的故障诊断的设计中,比较单一的神经网络和遗传神经网络的训练误差曲线,得出遗传神经网络在训练速度和准确性上远远高于神经网络训练模型.

To design Remoting examining system of speed-down on GA and NN

Through to the speed reducer failure mode's analysis, uses the BP neural network establishment failure diagnosis model, the use genetic algorithm optimization neural network weight, the threshold value, the network architecture, will inherit the neural network model to apply in the long-distance speed reducer's failure diagnosis design, compared with the sole neural network and the heredity neural network's training curve of error, obtains the heredity neural network is higher than the neural network training model by far in the training speed and the accuracy.