CAJ | 학술논문

针对传统基于单一方法的航电故障诊断专家系统存在的数据爆炸、知识获取代价与扩展难度大、诊断效率低等问题，提出了一种基于FTA、BAM神经网络、BP神经网络方法联合的故障诊断专家系统模型，模型保留了FTA对各种故障事件优秀的分析处理能力，发挥了BAM神经网络对单一或多源故障的快速定位及BP神经网络处理电路故障的优势，三者结合能够准确地找出故障位置和原因。利用VS2010平台研发了某型飞机航电故障诊断专家系统，以航电系统典型故障为诊断实例，验证了该专家系统诊断速度快、诊断结果准确，解决了传统专家系统存在的问题。
침대전통기우단일방법적항전고장진단전가계통존재적수거폭작、지식획취대개여확전난도대、진단효솔저등문제，제출료일충기우FTA、BAM신경망락、BP신경망락방법연합적고장진단전가계통모형，모형보류료FTA대각충고장사건우수적분석처리능력，발휘료BAM신경망락대단일혹다원고장적쾌속정위급BP신경망락처리전로고장적우세，삼자결합능구준학지조출고장위치화원인。이용VS2010평태연발료모형비궤항전고장진단전가계통，이항전계통전형고장위진단실례，험증료해전가계통진단속도쾌、진단결과준학，해결료전통전가계통존재적문제。
In the view of the traditional avionics expert diagnosis system based on single method,which exist problems such as data explosion,knowledge acquisition cost and difficult extension and low diagnosis efficiency,it puts forward a model of expert diagnosis system which based on the FTA,BAM and BP neu-ral network.This model retains the excellent ability of FTA which can deal with vorious failure event.It develop the advantages of BAM and BP neural network which can rapid locate the single and multiple source fault and process circuit malfunction.It can find out the fault position and cause accurately when comebine them together .Using VS2010 platform develops a certain type of aircraft avionic fault diagnosis expert system.Taking the typical fault of avionics system as a diagnosis example,which verify the expert system .It can diagnoe fast and accurate,which solves the problems existing in the traditional expert sys-tem.