论文部分内容阅读
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。
Aiming at the problems existing in the traditional avionics fault diagnosis expert system based on single method, such as data explosion, difficulty and cost of knowledge acquisition, and low diagnostic efficiency, a fault diagnosis based on FTA, BAM neural network and BP neural network is proposed Expert system model and model retain the excellent analysis and processing ability of FTA for various fault events and make use of the advantages of BAM neural network in quickly locating single or multi-source faults and BP neural network processing circuit faults. The combination of the three can accurately find The location and cause of failure. The VS2010 platform was used to develop an avionics fault diagnosis expert system for a certain type of aircraft. The typical fault of avionics system was used as a diagnostic example. The results show that the expert system has the advantages of fast diagnosis, accurate diagnosis and solved the problems existing in the traditional expert system.