基于BP神经网络的齿轮箱故障诊断

Fault Diagnosis of Gear Box Based on BP Neural Network

对BP神经网络的结构与原理进行了简要概述,将BP神经网络技术运用于齿轮箱的故障诊断中,以齿轮振动信号的时域特征作为神经网络输入,齿轮的主要故障形式为网络输出,利用经BP算法训练后的该网络对齿轮故障进行诊断,取得了较好的效果。

This paper briefly introduced the structure and principle of BP neural network,and applied the neural network technology to gear box fault diagnosis. The diagnosis system took the time domain characteristic parameters of gear vibration signal as the input of the network,and the gear failure form as the output. After BP algorithm training,the neural network was used to carry on the fault dignosis of gear box. The effect is satisfied.