基于神经网络的火控系统故障诊断专家系统  

Aviation Fire Control System Fault Diagnosis Expert System based on Neural Network

作　　者：程洪书[1,2] 孙卫[1] 蔡凯[1] 

机构地区：[1]空军工程大学工程学院,陕西西安710038 [2]空军军训器材研究所,北京100095

出　　处：《火力与指挥控制》2009年第1期120-122,共3页Fire Control & Command Control

基　　金：空军创新基金资助项目(CXJJ0521)

摘　　要：针对某型战机航空火控系统故障诊断方法对维修人员、检测设备依赖性大、故障诊断时间长等弊端,提出将神经网络与专家系统技术相结合应用于该型战机火控系统的故障诊断中。根据火控系统的故障特点给出了其神经网络专家系统的结构,建立了系统知识库,设计了系统运行的主界面和四大运行模块。经部队外场维修人员使用证明,该系统易于维护和使用,故障诊断准确率较高,有较强的鲁棒性。Aimed at that some fault diagnosis methods of aviation fire control system about a certain type of fighter airplane in existence have some disadvantages ,such as they are dependent on the maintainer and the detection equipment ,the time of fault diagnosis is too long and so on, the paper puts forward that neural network and expert system are combined to apply in the fault diagnosis of the fire control system. The structure of the neural network and expert system of fire control system are given according to their fault characteristics. At last the knowledge database of the system is founded,the system＇main face and the four important modules are designed. On the use of the outfield maintainer in the army it can prove that the system is easily maintained and used. The accuracy rate of the fault diagnosis is higher ,and the stability of the system is stronger too.

关 键 词：神经网络 专家系统 航空火控系统 故障诊断 

分 类 号：TP183[自动化与计算机技术—控制理论与控制工程]