论文部分内容阅读
针对机载燃油系统在线实时健康管理存在的观测信息不确定性、任务完成时限性的问题,研究了国内外最新健康管理算法,提出了机载燃油系统健康模型构建、在线实时推理的方法.该方法基于系统结构模型,采用面向对象方法构建BN(Bayesian Network)健康模型,并利用GVE(Global Variable Elimination)算法离线编译BN健康模型,构造AC(Arithmetic Circuit)健康模型.仿真结果表明:与BN健康模型相比,所设计的AC健康模型在观测信息存在不确定性的情况下,能够高精度在线诊断系统故障,也可以有效满足健康管理严格时限性要求.
Aiming at the uncertainty of observational information and the time limit of the task fulfilled in online real-time health management of on-board fuel system, the latest health management algorithms at home and abroad are studied, and a method of on-board fuel system health model construction and online real-time reasoning is proposed. Methods Based on the system structure model, a Bayesian Network (BN) health model was constructed by object-oriented method and the BN health model was compiled offline by GVE (Global Variable Elimination) algorithm to construct an AC (Arithmetic Circuit) health model.The simulation results showed that: Compared with the model, the designed AC health model can diagnose the system fault on-line with high precision under the uncertainty of the observed information, and also can effectively meet the strict time-limit requirements of health management.