摘要
对Petri网理论在性能评价领域的应用进行了研究,在介绍广义随机Petri网(GSPN,generalized stochastic petri net)及其分析技术的基础上,探讨了如何利用GSPN对系统进行建模及性能分析。针对典型的双处理机系统的吞吐量和资源利用率问题,建立了双处理机系统的GSPN模型,利用随机Petri网与马尔可夫链的同构关系,得到了GSPN模型稳定的状态空间及各状态出现的概率,进而探讨了双处理机系统的吞吐量和利用率等性能指标。通过TimeNet工具对所建立的模型进行仿真实验,验证了模型的有效性,为吞吐量和利用率的提高提供相关依据。
The application of Petri net theory to the field of performance evaluation is studied. Based on the generalized stochastic Petri nets and its analysis technique, utilization of the GSPN on system modeling and performance analysis is discussed. To solve the problem of throughput and resource utilization in typical dual processor system, a GSPN model of dual processor system is established, using isomorphic relations between stochastic Petri net and Markov chain. The stable state space of GSPN model and probability of each status are gained. Then the dual processor performance metrics such as throughput and utilization of the system are analyzed. Through the TimeNet simulation experiment tool, the validity of the established model is proved, which provides related basis for the throughput and availability improvement.
出处
《北京信息科技大学学报（自然科学版）》
2013年第6期38-41,共4页
Journal of Beijing Information Science and Technology University
基金
国家自然科学基金面上项目(61370065)
