基于Petri网矩阵模型仿真的改进算法  

New Simulation Algorithms Based on Matrix Model of Petri Nets

作　　者：李建军[1] 卫军胡[1] 赵健鸣[1] 孙国基[1] 

机构地区：[1]西安交通大学系统工程研究所,机械制造系统工程国家重点实验室,陕西西安710049

出　　处：《系统仿真学报》2005年第8期2015-2017,2021,共4页Journal of System Simulation

基　　金：国家863计划(2004AA412050)

摘　　要：Petri网的矩阵模型是分析Petri网的可达性、活性、有界性等基本性质以及状态演化过程的一种数学方法。在包含瞬时活动的情况下,传统的矩阵模型方法无法在一个步长内捕获所有应该同时激发的变迁,而是推迟到下一个步长,从而产生不精确的仿真结果。对此提出一种新的定步长仿真算法,得到了完全符合系统真实行为的仿真结果。算法进一步推广到基于事件调度的仿真方法中,可显著提高仿真效率。The matrix model of the Petri nets is a widely used mathematical tool for analysis of the properties of Petri nets,such as reachability, liveness, boundedness and etc. It is also used to explore the behaviors and the state changes of the simulated systems. In the situations where there exist transient activities, the traditional matrix model cannot identify in one time step all the transition firings that should happen at the same time. Instead, some of the firings are postponed to the next step. This therefore results in inaccurate simulation results. A new time-stepped simulation algorithm is presented based on the matrix model. The simulation results completely reflect the actual behaviors of the systems. The algorithm is further extended to the simulation implemented using the approach of next event scheduling.

关 键 词：离散事件系统 赋时PETRI网 矩阵模型 事件调度法 

分 类 号：TP391.9[自动化与计算机技术—计算机应用技术]