基于模糊贝叶斯网络的堆垛机系统可靠性分析  

Reliability analysis of stacker system based on fuzzy Bayesian network

作　　者：吕婷[1] 杨涛[1] 韩宾[1] 张宝[2] 

机构地区：[1]西南科技大学信息工程学院特殊环境机器人技术四川省重点实验室,四川绵阳621010 [2]四川烟草工业有限责任公司绵阳分厂,四川绵阳621000

出　　处：《计算机应用研究》2014年第12期3632-3636,共5页Application Research of Computers

基　　金：国家自然科学基金资助项目(11076024);四川省科技厅应用基础研究基金资助项目(2012JYZ003);川渝烟工技研【2013】165号

摘　　要：为解决因缺乏实际数据而无法准确估计堆垛机系统和部件的失效概率问题,提出了基于模糊集理论和主观贝叶斯方法的模糊贝叶斯网络诊断策略。该方法首先将故障树转换成相应的贝叶斯网络,然后运用模糊集理论,将专家给出的关于基本事件失效概率的主观语言评判值转换成模糊数,并通过去模糊化处理得到精确解。针对因事件的多态性所引起的条件概率不确定问题,该方法采用主观贝叶斯方法进行估计。通过堆垛机通信模块的可靠性分析实例,验证了该方法是有效的,表明其能够克服在系统建模时的参数不确定问题。In order to address the issue of estimating the failure probabilities of stacker system and its components accurately in the absence of real data,this paper proposed a diagnosis method that combined fuzzy set theory and subjective Bayesian theory with fuzzy Bayesian networks. This method firstly converted fault tree to the corresponding Bayesian network. Then it employed fuzzy set theory to map experts’ linguistic judgments on basic events’ failure probabilities into fuzzy numbers,defuzzified them and obtained precise results. It used subjective Bayesian theory to deal with the uncertainty of conditional probabilities caused by events’ multi-state. In the end this paper took the reliability analysis of stacker’s communication module as an example. The results show that the method is valid,it can overcome the uncertainty of parameters when modeling systems.

关 键 词：贝叶斯网络 模糊集理论 主观贝叶斯理论 不确定性 多状态 堆垛机通信模块 

分 类 号：TP183[自动化与计算机技术—控制理论与控制工程] TP202.1[自动化与计算机技术—控制科学与工程]