用D-S证据理论处理管道失效概率计算中的专家信息  

Using D-S Evidence Theory to Analyze Expert Information in Pipeline Failure Probability Calculation

作　　者：张鹏[1] 陈智勇[1] 彭星煜[1] 

机构地区：[1]西南石油大学建筑工程学院

出　　处：《天然气技术》2008年第1期44-47,共4页NATURAL GAS TECHNOLOGY

基　　金：国家自然科学基金(50678154);高等学校博士学科点专项科研基金(20060615003);教育部优秀青年教师资助计划项目;油气藏地质及开发工程国家重点实验室开放基金(PLN0113);中国石油科技风险创新研究项目(050511-5-2);四川省教育厅自然科学重点项目(2006A121)

摘　　要：现有的模糊故障树方法只聘请一位专家(或采用专家组综合意见)进行判断,未能集中多位专家的集体智慧,或未对专家组中个体信息进行精细分析。D-S证据理论可为多位专家的信息融合奠定理论基础,实现专家个体信息的去粗取精、去伪存真,使专家组的综合判断结果更接近真实情况。在故障树模糊分析法中引入D-S理论,不仅完善了管道失效概率的计算模型,更重要的是探索了在不考虑失效模式的情况下通过主观评判确定失效概率值的不确定信息处理方法。The existing fault tree analysis (FTA) just analyzes one expert advices to draw a conclusion, it could not collect many expert’s collectivity intelligence or analyze individual information of expert exactly. The D-S evidence theory can establish the theoretical foundation for fusing the data of many experts and achieve the prime and right individual data of experts to make the result judged by experts to be closer to the true. By this way it can not only make the calculating model of FTA to compute the failure probability perfect, but also bring forward the method to deal with uncertainty information such as ensuring the failure probability by judge of the expert when the failure model is scarcity.

关 键 词：管道 失效概率 模糊故障树分析 D-S证据理论 信任函数 置信度 

分 类 号：TP182[自动化与计算机技术—控制理论与控制工程]