航电设备故障诊断中的决策算法  

Algorithm of decision making for avionic device fault diagnosis

作　　者：李洁[1] 沈士团[1] 孙宝江[1] 齐怡[1] 

机构地区：[1]北京航空航天大学电子信息工程学院,北京100083

出　　处：《北京航空航天大学学报》2007年第6期677-681,共5页Journal of Beijing University of Aeronautics and Astronautics

基　　金："十一五"国防预研项目(B0520060455)

摘　　要：设计了故障诊断专家系统中的多属性决策算法.分析了诊断决策算法对诊断效率的影响.根据各属性对诊断决策的影响,对属性值进行规范化.采用线性加权函数来评价各故障的优先级,利用对故障发生概率的偏好性求解各属性权值.设计修正向量对评价函数进行修正以满足它在属性边界点上跳变特性.通过实验分析了该多属性决策算法在不同情况下对诊断效率的提高程度.The multiple attribute decision making（MADM） algorithm for fault diagnosis expert system was put forward. The influence of diagnosis decision-making on the diagnosis efficiency was analyzed. According to the effect of every attribute on diagnosis decision-making,the attributes＇ value was normalized. The linear additive weighting function was used to evaluate every fault priority. The weight value was determined by the preference information on the probability of fault occurrence. The approach of solving evaluation function correction vector was proposed to satisfy the jump characteristics on attribute boundary points. The enhancement of diagnosis efficiency under different conditions for the MADM algorithm proposed was strengthened by experimental analysis.

关 键 词：专家系统 多属性决策 故障诊断 

分 类 号：O212.5[理学—概率论与数理统计] V241[理学—数学]