基于信息融合技术的装备BIT故障诊断系统应用研究

BIT就是系统、设备内部提供的检测、隔离故障的自动测试能力;传统BIT主要面向单一功能设计,不具有同时应对多信息的处理功能,面对复杂电子装备的多信息处理问题,传统BIT故障诊断能力不足,导致系统虚警率过高,测试性与故障诊断能力下降,在BIT系统中运用信息融合技术对多信息进行整合,对于提高BIT综合效能,实现快速高效的故障诊断具有重要意义;文章首先对信息融合基本原理进行了介绍,构建了基于信息融合的装备BIT故障诊断系统结构,接着对信息融合诊断系统进行了建模,并对模型中基于神经网络的BIT诊断技术、基于D—S理论的BIT决策技术的关键技术进行了研究,最后给出了故障诊断程序流程;该技术研究通用性强,适用于多种不同型号的复杂电子装备的BIT快速故障诊断。

Research on the Application of Fault Diagnosis System Based on Information Fusion Technology for Equipment BIT

BIT(built-in test)is the automatic testing ability of fault-detection and fault-isolation which is provided by the inside of system and equipment.Traditional BIT is mostly designed towards the single function,but doesn't possess the capability of processing multi -information at the same time.Facing the problems of multi-information processing in complex electromechanical equipment,traditional BIT's fault diagnosis function is insufficient,which leads to high false alarm rate,and reduces the testability and the capability of fault diagno- sis.It is very important to apply information fusion technology in BIT for integrating multi-information,and it can improve the synthesis ef- fectiveness of BIT and realize quick and high efficiency fault diagnosis.Firstly,the paper introduces the basic principle of information fusion, and constructs a system structure of fault diagnosis based on information fusion technology for equipment BIT.Secondly,it establishes a model of diagnosis system based on information fusion and studies the key techniques of BIT diagnosis based on neural network and BIT deci- sion-making based on D-S theory of evidence in the model.At last,the paper presents a flow of fault diagnosis process.The research can be used widely.And it is fit for quick BIT fault diagnosis for many different kinds of complex electromechanical equipments.