为研究复杂系统的保障可靠性,在对保障系统任务可靠性、任务可靠度和任务可靠度密度及其估计进行定义的基础上,用PERT网络对保障任务流程进行建模,并用Monte-Carlo方法对PERT网络的任务工期进行仿真,进而进行基于仿真结果的任务工期概率特性分析、保障系统任务可靠性分析、路径及工作的重要度分析等.在对仿真结果的分析过程中,考虑工作持续时间服从非对称分布形式,选用比常用的正态分布形状适应度更好的[分布对任务持续时间进行概率密度函数拟合,并用粒子群算法对β分布参数进行寻优.对维修计划网络的仿真和分析表明,用Monte-Carlo方法对保障系统进行任务可靠性分析不受工作持续时间概率特性和工作间逻辑关系的限制,具有解析方法所不具有的广泛适用性,且基于粒子群的β分布参数寻优能以较高的精度估计出任务可靠性曲线的相关参数.%To research the supporting reliability of complex system, the mission reliability is defined and PERT network is used to model the process of supporting mission. In considering that the duration of the works are random variables subjected to β distribution, used Monte-Carlo simulation method to simulate the duration of stochastic PERT network, and analyzed the result of simulation comparatively. During the analysis, PSO used to estimate the β distribution parameters, and then the probability characteristics of the network' s duration, the mission reliability of the support system, analyzed the importance of line and work to mission reliability. The simulation and analysis of an equipment maintenance PERT network shows that the use of Monte-Carlo method to simulate the task duration and analysis the task reliability of supporting system is not restricted to the probability characteristics of the works' durations and the logical relationship between the works, so its applicability is broader than the analytical methods. In addition, the parameters of task reliability curve can be estimated with high accuracy by particle swarm optimizer.
展开▼
