一种故障可诊断性分析方法

A Method for Analysing Sensitivity of Fault

介绍了应用结构排除法进行故障诊断时故障的可诊断性,指出了使用(0,1)故障矩阵时有可能出现错误的诊断结果,介绍了计算敏感度矩阵的方法,同时也和(0,1)故障矩阵进行了比较。在设计故障诊断系统时,利用敏感度矩阵,能够针对不同的关系式故障选择最佳检验参数和排除参数的组合,可以提高故障诊断结果的可信度。

This paper discusses the sensitivity of fault in fault diagnosis when structuralelimination method is applied.Fault diagnosis in the structural elimination is based on mathematical model and measured parameters.Because the measured parameters vary with the different faults,it is necessary to make deeper analysis when employing (0,1) fault matrix to reduce the possibility of making wrong diagnosis.The analytical method mentioned above which uses the sensitivity matrix is proposed in this paper.The method of calculating sensitivity matrix and an example that uses the matrix are also presented.The sensitivity matrix as well as the sensitivity degree of parameters for fault is compared with (0,1) fault matrix.The best test parameters can be chosen and the parameter's combination can be eliminated according to different relationships of fault when the fault diagnosis system is designed by using the sensitivity matrix.Therefore,the reliability of the result of fault diagnosis will be improved.