基于贝叶斯网络的相控阵雷达模糊故障诊断系统设计

Design of Phased Array Radar Fuzzy Fault Diagnosis System Based on Bayesian Networks

机载有源相控阵雷达由于内部数字化器件集成度高、技术复杂等因素,其BIT测试点设置和设计难度很大,单纯依赖机上BIT进行故障诊断很难准确隔离到单个LRM。该文在分析某型号雷达BIT诊断故障模糊特点的基础上,设计了一种基于贝叶斯网络的故障诊断系统,利用专家知识构建网络结构,通过EM算法确定数据不完备条件下的网络参数,采用联结树算法实现故障诊断推理,从而提高了模糊故障的诊断精度和排查效率。最后,通过实例验证了该系统的有效性和可行性。

Due to high integration and complex technology of the airborne active phased array radar internal digital devices,it is hard to set up and design test points of the radar,relying solely on BIT for fault diagnosis is difficult to accurately isolating to one LRM.After analyzing the fuzzy characteristics of a radar BIT fault diagnosis,the fault diagnosis system based on Bayesian networks is designed,using expert knowledge to construct the structure,determining the parameters under the condition of incomplete data by EM algorithm,realizing fault diagnosis reasoning through junction tree algorithm.The accuracy and efficiency of fuzzy fault diagnosis are improved.Finally,the effectiveness and feasibility of the fault diagnosis system are proved with examples.