考虑时间相关故障的多状态系统可靠性与任务成功性仿真评估方法  

Simulation assessment method for multi-state system reliability and mission success probability considering time-dependent faults

作　　者：杨皓洁 吕建伟[1] 徐一帆 YANG Haojie;LYU Jianwei;XU Yifan(Department of Management Engineering and Equipment Economics, Naval University of Engineering, Wuhan 430033, China)

机构地区：[1]海军工程大学管理工程与装备经济系,湖北武汉430033

出　　处：《系统工程与电子技术》2021年第8期2362-2372,共11页Systems Engineering and Electronics

基　　金：国家自然科学基金(7140117);军委装备发展部装备预研领域基金(6140002050116JB11002);中国博士后科学基金(2019M653925);军队院校2110Ⅲ期基金项目(4142D45A3)资助课题。

摘　　要：针对考虑时间相关故障、多状态性能输出、系统冗余储备等因素的系统可靠性与任务成功性评估问题使用解析法难以建模求解,而一般仿真方法又存在抽样效率慢和估值精度低的问题,提出具有时间相关故障抽样与统计方差减小相结合的蒙特卡罗仿真评估方法。首先,设计了非指数剩余分布抽样的数值计算方法,实现了时间相关故障的随机事件抽样。其次,在此基础上结合受迫转移、失效偏倚等方差减小方法,提高了小概率事件的仿真抽样效率,改善了仿真评估方法的估值精度。最后,通过舰船航渡任务的算例,验证了该方法对评估复杂系统行为和故障机制条件下系统可靠性及任务成功性的合理性和有效性。Aiming at the problem of system reliability and mission success probability assessment considering time-dependent faults,multi-state performance output,and system redundancy,the analytical method is difficult to model and solve,and the general simulation method has poor sampling efficiency and low estimation accuracy.A Monte-Carlo simulation assessment method combining time-dependent faults sampling and statistical variance reduction is proposed.First,a numerical algorithm for non-exponential residual distribution sampling is designed to implement random event sampling for time-dependent faults.Then,combining with the variance reduction methods such as forced transition and failure biasing,the simulation sampling efficiency of small probability events and the estimation accuracy of the simulation method are improved.Finally,the example of the warship’s sail mission verifies the rationality and effectiveness of the proposed method for assessing the system reliability and mission success probability under complex system behavior and fault mechanism.

关 键 词：任务可靠性 多状态系统 非指数分布 蒙特卡罗间接仿真法 方差减小方法 

分 类 号：U674.7[交通运输工程—船舶及航道工程] TB114.3[交通运输工程—船舶与海洋工程]