A Problem Solver for Scheduling Workflows in Multi-agents Systems Based on Petri Nets

Abstract

Workflow scheduling in multi-agent systems (MAS) is a challenging problem due to the computational complexity involved, distributed architecture and dependency of different agents' workflows. How to develop a problem solver that can be applied in MAS to achieve coherent and consistent workflow schedules that can meet a customer's order is an important issue. In this paper, we propose a solution methodology for scheduling workflows in MAS. Our solution combines the multi-agent system architecture, contract net protocol and workflow models specified by Petri nets. Our solution algorithm is developed based on transformation of workflow model to network models. A subgradient algorithm and a heuristic algorithm are applied to find the solutions. A problem solver for workflow scheduling in MAS has been implemented.