Hybrid heuristic search for the scheduling of flexible manufacturing systems using Petri nets

The combination of Petri nets (PNs) as an analysis tool for discrete-event dynamic systems and artificial intelligence heuristic search has been shown to be a promising way to solve flexible manufacturing systems (FMS) scheduling problems. However, the NP hard nature of the problem obscures the PN capability of reasoning about the behavior of the system. In this paper, two techniques to alleviate this drawback are presented: a systematic method to avoid the generation of futile paths within the search graph and a novel hybrid stage-search algorithm, The new algorithm is based on the application of A* guided by a PN-based heuristic within a limited local search frame. An optimization policy is applied to maintain, under evaluation, only the most promising paths. For each system state, the algorithm is able to decide whether an enabled operation should be applied and to maintain this decision until new information forces reconsideration. This eliminates permutation paths and useless scheduling sequences. Experimental results show that the algorithm's cost does not grow exponentially with the size of the problem. Comparison with previous work is given to show the superiority of our approach and the potential of PN-based heuristic search.