基于拆卸及扩展广义随机 Petri网的维修性仿真

Maintainability Simulation Based on Disassembly and Extended Generalized Stochastic Petri Nets

摘要:

在装备研制早期,难以有效地利用试验的方法进行维修性估计,此时适合采用仿真手段。对于机械系统拆卸路径的分析与优化,先基于赋时变迁 Petri网建立拆卸 Petri网模型,再利用 A*算法搜索最优拆卸路径,并确定拆卸时间。为了建立系统整个维修事件的过程模型,对广义随机 Petri网进行了扩展,将其延时变迁与实际分布关联,对基于变迁的维修作业之间的逻辑关系进行了描述,并给出了基于蒙特卡罗法的维修时间仿真算法。最后,以对某火炮装备发射系统不击发故障的分析与排除为例,仿真计算了维修时间均值与方差,并对结果进行了分析。

Abstract:

In the equipment early development, it is hard to do maintainability estimation effectively by means of test, but the simulation method is appropriate. As to the disassembly process’s analysis and optimization, the dis. assembly Petri net model was set up based on timed transition Petri net first, then the A* algorithm was used to search the optimal disassembly path and measure disassembly time. To establish the whole maintenance event process model, the model of generalized stochastic Petri net was extended so that its timed transitions could be associated with actual distributions, the timed transition based maintenance activities and their logic associations were described; and the maintenance time simulation algorithm based on the Monte Carlo method were given. Last, aiming at the analysis and obviation of some gun’s emission system failure to fire, the mean and variance of maintenance time were calculated by simulation, and the results were analyzed.