摘要
为提高装备维护水平,立足于某型潜用火控雷达,着眼于海军装备故障诊断现状,设计了一种基于多Agent的智能故障诊断系统总体结构,诊断系统能根据设备的结构变化进行动态调整,通过知识库和诊断推理机的分离、诊断Agent的动态生成和管理等技术,使该智能故障诊断系统能方便地应用到其它装备。以BDI结构模型为基础,从基于实例的推理(CBR)、产生式诊断系统和神经网络诊断系统三个方面实现了诊断Agent的诊断功能,可最终将故障定位到可更换的元器件。
With a view to the actuality of navy equipments fault diagnose,the architecture of intelligent diagnostic is designed based on Multi-Agent（MAS） according to some sort of radar on submarine to improve the ability of equipments maintenance.The diagnostic system is adjusted according to the structure of the equipment.By setting apart the knowledge base and the reasoning machine,creating and managing the diagnostic agent dynamically,this diagnostic system can be generalized to be used for other equipments.Based on BDI model,the main function of diagnostic agent is implemented in three aspects as CBR（Case Based Reasoning）,production diagnostic system and artificial neural network.The fault will gradually be located at the replaceable component level.
出处
《舰船电子工程》
2013年第3期110-113,共4页
Ship Electronic Engineering
