Heuristic search and concurrency strategy based on Petri net for FMS scheduling

This paper proposes a concurrency strategy based on timed-transition Petri net for scheduling flexible manufacturing system (FMS). Combined with the heuristic function, prior search algorithms explore only one transition during each execution, which brings in many redundancy states for the expending procedure. In this paper, we prove that firing some non-conflicting transitions simultaneously can also achieve an optimal schedule. This can help the search algorithm explore deeper successor states and avoid many unnecessary intermediate states. As a result, the search effort can be reduced apparently while the optimality of result is still guaranteed. The experimental results show that the proposed concurrency strategy performs better than the reported methods.