Efficient synthesis of production schedules by optimization of timed automata

For industrial systems with multiple products or alternative production paths, the time-optimal assignment of production steps to available units is often economically indispensable. Among the several methods proposed for solving such job-shop scheduling problems, techniques based on reachability analysis for timed automata (TA) have gained attention recently. It was proposed recently to integrate techniques known from mixed-integer programming (MIP) into reachability analysis for TA to establish an efficient pruning of the search graph. This paper further improves the efficiency of the algorithm by excluding suboptimal or redundant solutions from the search space. For several benchmark problems, it is shown that the proposed approach can produce better schedules in a given computation time as are obtained by applying pure MIP to large problem instances.