基于BP神经网络的某型飞机火控系统故障诊断

Fault diagnosis of a certain type of aircraft fire control system based on BP neural network

根据某型飞机火控系统的特点,给出了BP神经网络数学模型及其学习算法,在此基础上以某型飞机火控系统作为为被诊断对象,运用BP神经网络数学模型及其学习算法对其进行故障诊断。诊断结果表明BP神经网络不仅能识别出样本自身的故障,而且能准确诊断出样本以外数据故障,提高某型飞机火控系统一线维修保障效率。

According to the characteristics of the fire control system （FCS） on a certain type of aircraft, we select Error Back Propagation （BP） Network for fault diagnosis of the FCS. The mathematic model and leaming method of BP Network were studied. Then we used the model and the learning method for fault diagnosis of the FCS. The diagnosis result indicated that the BP neural network can not only identify effectively the fault of the sample itself but also identify accurately the fault out of the sample, thus the field maintenance efficiency of the FCS can be improved.