基于小波能谱熵和HMM的机载雷达故障预警技术  

Fault Early Warning Technology of Airborne Radar Based on Wavelet Energy Spectrum Entropy and HMM

作　　者：陈建平 徐皓吉 吴庭金 CHEN Jianping;XU Haoji;WU Tingjin(AVIC Leihua Electronic Technology Research Institute,Wuxi 214063,China)

机构地区：[1]中国航空工业集团公司雷华电子技术研究所,江苏无锡214063

出　　处：《火力与指挥控制》2022年第10期31-35,40,共6页Fire Control & Command Control

基　　金：航空科学基金资助项目(20172007002)。

摘　　要：机载雷达服役中常出现性能退化、间歇故障等传统BIT无法识别的故障征兆,工程上迫切需要故障预警技术。研究了考虑退化与间歇故障的预警技术。阐明检测原理,提出测试指标集;采用NDGM模型描述小波能谱熵的增长情况;建立基于HMM和正态云模型的故障预警模型,给出参数设计方法,建立故障预警决策流程。在案例研究中,得到组件状态及剩余寿命,给出预防性维护结论,表明所提方法可解决机载雷达故障预警问题。In the service process of airborne radar,such fault symptoms as performance degradation and intermittent failure and others often occur,which cannot be identified by traditional built-in test(BIT)methods.Therefore,the demand for fault early warning in engineering is urgent.To solve this problem,the early warning technology considering degradation and intermittent fault is studied.Firstly,the detection principle is clarified and the test index set is proposed.Secondly,non-homogenous discrete grey model(NDGM)is adopted to describe the growth of wavelet energy spectrum entropy.Thirdly,a fault early warning model based on hidden Markov model(HMM)and normal cloud model is established,the design method of HMM parameters is given,and the fault early warning decision-making process is established.The component state and remaining life are obtained,and thus the preventive maintenance conclusion is given in the case study,which shows that the proposed method can solve the problem of airborne radar fault early warning.

关 键 词：故障预警 机载雷达 小波能谱熵 隐马尔可夫模型 正态云模型 

分 类 号：V240.2[航空宇航科学与技术—飞行器设计] V243.2