核主泵机械密封可靠性Bayes分析的适应性

Adaptability of Bayes Analysis for Reliability of Mechanical Seals of Nuclear Main Pumps

机械密封的可靠性对核主泵的安全稳定运行具有重要影响。为解决无失效数据情形下核主泵机械密封的可靠度评估问题,分析大亚湾核主泵机械密封的运行数据,确定其可靠度分布,建立结合Bayes理论的可靠性分析模型,利用Monte Carlo法从确定的可靠度分布中仿真出无失效数据,探讨无失效数据场合下,先验分布为Beta分布时,分组数c的取值对E-Bayes估计与多层Bayes估计精度的影响。研究表明:分组数c<8时,优先选择多层Bayes估计;c>8,优先选择E-Bayes估计,c=8时,2种方法的平均相对误差均达到较低水平且多层Bayes估计更低一些。研究成果对无失效数据场合下基于Bayes理论的核主泵机械密封可靠性分析具有指导意义。

The reliability of mechanical seal has an important influence on the safe and stable operation of nuclear main pump.To achieve the reliability evaluation of the mechanical seal of the nuclear main pump under the condition of zerofailure data,the operating data of the mechanical seal of the Daya Bay nuclear main pump were analyzed,its reliability distribution was determined.A reliability analysis model combining Bayes theory was established,and the zero-failure data was simulated from the determined reliability distribution by using Monte Carlo method.The influence of the value of packets c on the accuracy of E-Bayes estimation and multi-layer Bayes estimation was investigated when the prior distribution was Beta distribution in the case of zero-failure data.The results show that when packets c<8,multi-layer Bayes estimation is preferred,when c>8,E-Bayes estimation is preferred,and when c=8,the average relative errors of both methods reach a lower level and the multilayer Bayes estimation is lower.The research results have guiding significance for the reliability analysis of mechanical seal of nuclear main pump based on Bayes theory in the case of zero-failure data.