基于Bayes网络高性能滚动轴承失效影响因素可靠性评估方法  

Reliability Evaluation Method for Failure Factors of High Performance Rolling Bearing Based on Bayes Network

作　　者：韩兴[1] 李昌[1] 陈正威 李云飞 HAN Xing;LI Chang;CHEN Zheng-wei;LI Yun-fei(School of Mechaniccil Engineering and Automation,University of Science and Technology Liaoning,Liaoning Anshan 114051,China;Energy China NEPC,Liaoning Shenyang 110179,China)

机构地区：[1]辽宁科技大学机械工程与自动化学院,辽宁鞍山114051 [2]中国能建东电一公司,辽宁沈阳110179

出　　处：《机械设计与制造》2022年第2期171-176,共6页Machinery Design & Manufacture

基　　金：国家自然科学基金(E050402/51374127);辽宁省自然科学基金指导计划项目(2019ZD0277);辽宁省教育厅项目(2017FWDF01);辽宁科技大学创新团队建设项目(601009830)。

摘　　要：滚动轴承是精密机床、高速列车、航空发动机等重大装备的关键核心部件,其性能的好坏直接影响到主机的使役寿命。因此,对高性能滚动轴承失效形式、损伤机理进行相关影响因素可靠性定量评估具有重要意义。首先建立了高性能滚动轴承失效影响因素故障树模型,直接映射建立其Bayes网络模型,利用桶排除法对轴承失效概率及中间失效结点概率进行了计算,利用全概率公式对轴承失效过程各影响因素的条件概率进行可靠性定量评估。计算表明:接触疲劳剥落对轴承失效贡献概率最大,为15.454%,轴承零件内部夹杂物失效的条件概率为0.058%,影响最小。Rolling bearings are key components of precision machines,high-speed trains,aero engines and other major equipments,and their performances directly affect the host’s service life. Therefore,it is very important to evaluate the reliability of the high performance rolling bearing failure modes and the damage mechanisms. Firstly,a fault tree model of high performance rolling bearing failure factors was established,and its Bayesian network model was established by direct mapping. Secondly,with the barrel exclusive method,it calculated the bearing failure probabilities and intermediate failure nodes probabilities. Lastly,it used the full probability formula to reliability quantitative evaluation the conditional probability of each factor of the bearing failure process. The results showed that the contact fatigue flaking had the greatest influence on bearing failure,and was 15.454%;the conditional probability of failure of internal inclusions in bearing parts was 0.058%,and it had the least influence on bearing failure.

关 键 词：高性能滚动轴承 失效影响因素 Bayes网络法 桶排除法 可靠性定量评估 

分 类 号：TH16[机械工程—机械制造及自动化] TH122