摘要
An evaluation approach for the response time probability distribution of workflows based on the fluid stochastic Petri net formalism is presented. Firstly, some problems about stochastic workflow net modeling are discussed. Then how to convert a stochastic workflow net model into a fluid stochastic Petri net model is described. The response time distribution can be obtained directly upon the transient state solution of the fluid stochastic Petri net model. In the proposed approach, there are not any restrictions on the structure of workflow models, and the processing times of workflow tasks can be modeled by using arbitrary probability distributions. Large workflow models can be efficiently tackled by recursively using a net reduction technique.
提出了一种基于流体随机Petri网的工作流响应时间概率分布计算方法.首先讨论了利用随机工作流网建模的一些相关问题,然后描述了如何将随机工作流网模型转化为流体随机Petri网模型,最后给出了该种流体随机Petri网模型的动态方程,说明工作流响应时间的概率分布可直接由流体随机Petri网模型的暂态解得到.该方法对工作流模型的结构没有提出任何限制,且工作流任务的处理时间可取任意概率分布,通过递归地使用网化简技术可有效地处理大型工作流模型.
基金
The National Natural Science Foundation of China(No.60175027).
