失效概率计算中的信息遗失与系统级建模方法  

Information Loss in Failure Probability Calculation and System-level Modeling Approach

作　　者：谢里阳[1] 吴宁祥[1] 王正[1] 

机构地区：[1]东北大学,沈阳110004

出　　处：《中国机械工程》2007年第19期2277-2283,共7页China Mechanical Engineering

基　　金：国家重大基础研究计划资助项目(2006CB605000);国家863高技术研究发展计划资助项目(2006AA04Z408)

摘　　要：通过对系统失效概率问题的深入分析,揭示了传统上通过载荷-强度干涉分析计算零件失效概率过程中发生的系统失效相关性信息遗失问题,以及无法通过零件失效概率指标构建一般系统(具有失效相关性的非独立失效系统)失效概率模型的原因。从数学的角度探讨了载荷-强度作用机制和一般系统都存在共因失效的原因。在此基础上,提出了通过"系统级"的载荷-强度干涉关系建立系统失效概率模型的方法,建立了能根据有限的失效数据预测系统任意阶失效概率的参数化模型。最后,通过实例验证了所提出模型的适用性。Through a thorough analysis to the essence of the method to calculate component failure probability via load-strength interference relationship, the phenomenon of system failure attribute related information loss was revealed, and the reason why it was impossible to establish s-dependent system failure probability model simply by means of component failure probabilities was interpreted. This paper also discussed the mechanism of common cause failure （CCF） for general system. Based on system-level load-strength interference analysis, the attribute of component failure probability was presented and the relationship between component failure probability attribute and load/ strength distribution was described. The coupling mechanism for common cause failure was explained mathematically. A parametric model was developed to estimate system failure probability according to limited amount of failure event data available. Examples show that the proposed model has a good consistency with experimental results.

关 键 词：系统失效概率 干涉理论 信息遗失 表决系统 共因失效 

分 类 号：TB114.3[理学—概率论与数理统计]