| Workflow technology is the core technology which realizes enterprise process modeling, simulation analysis, optimization, management and integration, and ultimately automation. More and more enterprises are widely using workflow technology to improve operation efficiency. Time management is a very important issue when workflow technology is used in enterprises, which will directly influence workflow efficiency. In general, resources are limited in enterprises. When implementing workflow management system, the resource constraint may influence workflow time management, including deadline, time performance, etc.In order to solve the problems above, this paper first proposes a method of time verification with resource constraint. The method uses workflow graph to model workflow and judge whether workflow instances executing concurrently will violate their deadlines or not when competing resources through calculating workflow critical path, which provides important information for workflow system. Workflow engine can schedule the workflow instances which compete resources according to relevant rules in order that as many as possible workflow instances can satisfy their deadlines.Workflow performance analysis is a very important aspect in workflow research, which can give important parameters of temporal characteristic. Workflow designers and administrators can use these parameters to improve workflow operation efficiency. The paper presents a performance analysis method which uses General Stochastic Petri Nets to model workflow and applies queueing theory to the evaluation of the influence on workflow performance caused by resource constraint. Sequence, choice, parallel and loop, four basic Stochastic Petri Nets model and their improved equivalent performance formula are used to calculate workflow performance parameters and several methods are also given to improve workflow performance. Finally, simulation tool ExSpect is used to verify theoretical analysis results. |
PDF Full Text Request