Scheduling of shipyard block erection system based on Petri net and GASA

To solve scheduling problem of shipyard block erection system (SBES), a scheduling method combining timed Petri net (TPN) and genetic algorithm simulated annealing (GASA) was proposed. The model based TPN is available to not only show the multiprocessing paths for erection process to be processed, but also depict resource constrains on each process. The definition of Petri net is extended to accord with the real-world SBES´s organization, and the modeling method for Erection Network Plan Graphic is provided in order to simplify modeling program. In addition, a hybrid algorithm which combined GA with simulated annealing (SA) was developed to find the optimal solutions of this problem. In the algorithm, the firing sequences of Petri net model was coded by welding transition and selecting places of the TPN model. Thus the selection, crossover and mutation operators deal with the elements of TPN instead of the ones of the problem space. Computational simulation suggests that the proposed TPN-GASA scheduler can provide obvious improvement comparing with conventional scheduling method.