多状态机械系统可靠性的离散化建模方法  被引量:13

Discretized Modeling Process of Reliability of Multi-state Mechanical Systems

在线阅读下载全文

作  者:钱文学[1] 尹晓伟[2] 谢里阳[1] 

机构地区:[1]东北大学机械工程与自动化学院,辽宁沈阳110004 [2]沈阳工程学院机械工程系,辽宁沈阳110136

出  处:《东北大学学报(自然科学版)》2008年第11期1609-1612,1632,共5页Journal of Northeastern University(Natural Science)

基  金:高等学校博士学科点专项科研基金资助项目(20070145083);国家高技术研究发展计划项目(2006AA04Z408)

摘  要:针对传统的基于二态逻辑的可靠性评估方法应用于多状态系统理论和实际应用存在差异的问题,根据贝叶斯信念网(BBN)具有双向不确定性推理功能和图形化显示的特点,提出了一种多状态机械系统可靠性离散化建模方法.首先确定BBN的节点及离散系统各元件的多个状态,并给出各状态的概率,用概率分布表(CPD)描述元件各状态之间的关系来表达关联节点的状态,最终建立离散化BBN模型.该模型避免了已有方法复杂的公式计算,对元件数量没有限制.实例分析表明了应用BBN离散化模型进行多状态机械系统可靠性评估的有效性和优越性.It was found that there is a discrepancy between theory and practical effect when applying the conventional reliability assessment based on binary logic to the multi-state mechanical systems, especially the result of its application is unsatisfactory. A new discretized modeling process is therefore proposed on BBN(Bayesian belief networks) basis for the reliability of multi-state mechanical systems, since BBN is featured with bidirectional reasoning function for uncertainty and graphic display. The discretized BBN model is developed the way the nodes in BBN and the multi-states of all components in discrete system should be determined to give the probability of each and every state, then the CPD(conditional probability distributing) is used to describe the relationships between different states so as to express the states of relevant nodes. In the model the complex computation due to various equations in earlier works can be avoided with the number of components unlimited. An example is given and the analyzing results show the effectiveness and favorableness of the discretized BBN model if using it to assess the reliability of a multi-state mechanical system.

关 键 词:可靠性 多状态系统 贝叶斯信念网 离散化 建模 

分 类 号:TH112[机械工程—机械设计及理论] V232.3[航空宇航科学与技术—航空宇航推进理论与工程]

 

参考文献:

正在载入数据...

 

二级参考文献:

正在载入数据...

 

耦合文献:

正在载入数据...

 

引证文献:

正在载入数据...

 

二级引证文献:

正在载入数据...

 

同被引文献:

正在载入数据...

 

相关期刊文献:

正在载入数据...

相关的主题
相关的作者对象
相关的机构对象