应用BP神经网络诊断全液压坑道钻机故障

Application of BP neural network in fault diagnosis of full hydraulic drilling rig

将BP神经网络原理应用于全液压坑道钻机现场故障诊断。根据全液压坑道钻机的故障特点,选取能够表征全液压钻机状态的量,采用归一化的方法确定样本的输入和输出;通过对网络进行训练,获得具有一定泛化能力的网络;利用Matlab软件实现此种故障诊断过程,获得了正确的诊断结果。试验表明,该法有一定的实用性,对于建立故障自动诊断系统有重要意义。

Choosing status messages according to fault diagnosis＇s characteristic, BP neural network is used in fault diagnosis of full hydraulic drilling rig. The sample input and output are used in unitary method and a broad neural network is obtained through training network. It is very important to construct fault self-diagnosis using Matlab software, and the correct result is gained. Experiment indicated that this method is practical and has significant importance for construction of automatic fault diagnosis system.