机载燃油系统在线实时健康管理

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.

Online real-time health management for aerial fuel delivery system

To address the problems of observational information uncertainties and task deadlines encountered in the online real-time health management(HM) for aerial fuel system(AFS), the latest HM algorithms in China and aboard were researched, and an approach for health modeling and online real-time reasoning of AFS was proposed. The object-oriented method was used to develop BN health model based on the system structure model, and the global variable elimination(GVE) algorithm was applied to the offline compilation of this BN into an arithmetic circuit(AC) health model. The simulation results show that the AC, compared with the Bayesian network, can not only provide precisely online diagnosis of system faults, but also effectively meet the strict time deadlines for HM requirements, under the condition of observational information uncertainties.