«上一篇/Previous Article|本期目录/Table of Contents|下一篇/Next Article»

《计算机技术与发展》[ISSN:1006-6977/CN:61-1281/TN]

卷:

24

期数:

2014年05期

页码:

1-4

栏目:

智能、算法、系统工程

出版日期:

2014-05-31

文章信息/Info

Title:

Research on Model Parameter Identification Based on Normal Stochastic Petri Net

文章编号:

1673-629X（2014）05-0001-04

作者:

陈哲; 刘久富; 王正谦

南京航空航天大学 自动化学院

Author(s):

CHEN Zhe; LIU Jiu-fu; WANG Zheng-qian

关键词:

非马尔可夫网; 随机Petri网; 模型设计; 参数识别

Keywords:

non-Markov nets; stochastic Petri nets; model design; parameter identification

分类号:

TP31

文献标志码:

A

摘要:

时间因素在对实时系统的分析中占据着重要的位置。随着时间因素在Petri网应用分析中的地位越发重要，各种与时间相关的Petri网模型相继被提出。文中结合动态实时系统的运行特征，设计了一种变迁点火时间服从正态分布的随机Petri网模型，并基于对系统行为的统计分析，提出了一种用于对变迁时间参数进行识别的数值算法。以一种随机制造单元为例，建立正态随机Petri网模型，并运用参数识别算法确定变迁时间参数。仿真结果符合系统的实际运行结果，证明了该方法的有效性。

Abstract:

The factor of time had played an important role in the analysis of real-time systems. As the importance of time analysis in the application of Petri nets is increasing,a variety of timed Petri net models were proposed. Considering the characteristics of dynamic real-time systems,a stochastic Petri net with normal distributions of the firing periods was designed and a method for the identification of the parameters was proposed based on a statistical analysis of the behavior of the system. The system model of a random manufacturing unit was built with normal Petri net and the parameters of firing periods were verified by the estimation algorithm. The simulation results fit in-to the actual behavior of the system and prove the effectiveness of this method.

常用功能

更新日期/Last Update: 1900-01-01