基于历史数据的测试任务约简和故障诊断

针对部队航电组件测试任务繁重、故障定位率低的问题，提出一种利用历史数据来简化测试任务、提高故障定位率的方法。利用粗糙集信息系统理论，建立了航电组件故障信息系统模型；基于测试任务辨识函数和诊断允许误差对测试任务进行约简；基于最短测试时间选出最优测试任务集；利用贝叶斯最大后验概率进行故障诊断推理，将诊断问题归结为不等式约束极值问题；用0-1规划隐数算法求得最优解。最后以某型飞机惯导部件为例验证了方法的快速有效性。

Test tasks reduction and fault diagnosis based on historical data

Aiming at the problem of complex test tasks and low fault location accuracy to avionics,a method based on historical data is proposed to reduce the test tasks and improve the fault location accuracy.A model of the avionics information system is built by using the rough set theory;the test tasks are reduced by using the novel discernibility function and minimum diagnosis error;the optimized test task set with minimum test time is chosen;the Bayes maximal posterior probability is introduced into fault diagnos...