一种改进的反向传播算法在机械故障模式识别中的应用

反向传播算法在机械故障模式识别中得到广泛的应用，但在实践中发现标准的反向传播法的收敛速度不快，模式识别能力有发，针对这些缺点进行改进，提出了一种改进的反向传播算法，大大提高了ＢＰ网络的收敛速度和模式识别能力，然后以变速箱故障模式识别为例进行了验证。

Applications of the Improved BP Algorithm to Machinery Fault Diagnosis

The algorithm of the back-propagation neural network has found wide applications in Machinery Fault Diagnosis. However, the generalization velocity and fault identification ability of BP algorithm are restricted in practice. A new improved BP algorithm which enhances the generalization velocity and fault identification ability is proposed in this paper. Finally, the author took fault diagnosis of gearbox as an example.