基于神经网络的某导弹交流地面电源故障诊断

Fault diagnosis for certain missile AC ground electrical source based on neural network

设计了一种基于神经网络的故障诊断系统,利用神经网络的独特优点,以实现对某导弹交流地面电源容易出现的故障进行快速定位和智能诊断。实验结果表明:神经网络不仅能够识别出所有已经训练过的模式,还可对未曾训练过的模式具有一定的联想推理能力。

A fault diagnosis system based on neural network is designed, and the particular merit of neural network is used to make quick orientation and intelligent diagnosis of easy-happening faults for certain missile AC ground electrical source. Result of the experiment indicates that neural network can not only identify all of the modes which have been trained, but also have certain ratiocination ability for the modes which have not been trained.