基于信息流模型的自动驾驶仪多故障诊断分析

Analysis of Multi-Fault Diagnosis for Autopilot Based on Information Flow Model

为了在多故障情况下对自动驾驶仪故障做出正确诊断,提出了基于相关性矩阵的多故障诊断完备性分析方法,并给出了严格的数学证明;首先对系统的软硬件模块进行分析与划分,利用信息流模型建立故障与测试间的相关性矩阵模型;再根据多故障情况下的诊断完备性分析方法对相关性矩阵进行行列变换,说明在多故障情况下此自动驾驶仪的故障诊断正确率理论上能达到100%,最后构建其故障诊断方法;实际应用表明,此方法对于自动驾驶仪的分析是有效的,可满足系统的指标要求。

In order to diagnose all faults of the missile autopilot correctly in multi-fault condition,an analytical method for multi-fault diagnosis completeness based on dependency matrix was proposed,and the strict mathematic proof was given.Firstly,the software and hardware of autopilot was analyzed and divided into different modules,and the information flow model was used to build the dependency matrix model between fault and test.Then the dependency matrix＇ s ranks were transformed based on the analytical method for muhi-fauh diagnosis completeness.It proved that the correct rate of fault diagnosis could reach 100％ theoretically in the case of multiple faults.Finally the diagnosis method for the missile autopilot was got.The results show that the method is correct and can meet the requirement of system index.