基于对向传播神经网络的齿轮箱故障诊断研究

Diagnosis to gearbox based on counter propagation neural network

介绍了CPN网络的原理、算法,并用CPN网络对齿轮箱机械传动系统的故障进行诊断。实例结果表明:该方法能够准确地诊断齿轮箱故障,具有训练速度快、结构简单、精度高等特点,是一种行之有效的诊断方法。

The paper introduces principles and algorithms of CPN network,and applies fault diagnosis to the mechanical transmission system of gearbox with the CPN network.The result from an example shows that the method can accurately diagnose the gearbox failure,features quick training speed,simple structure and high precision,and is an effective diagnosis method.