Batch Scheduling Algorithm For Mould Heat-treatment Two Stage Flow Shop

Heat-treatment operation is the bottleneck of the mould manufacturers. Effective control of heat-treatment production can not only increase the equipment utilization rate, shorten production cycle and improve the delivery performance, but also play an important role on saving energy, reducing emission and reducing cost in mould manufacturing. Quenching and tempering are two mainly processes in heat-treatment shop floor, which is a flow shop with batch machines. Assuming that the jobs differ from each other in scale, release date, due date and weight, this paper study the scheduling mechanism, optimization models and solution algorithms of the mould heat-treatment production.The research contents are divided into four modules from basic to complex. First of all, the mould heat-treatment shop flow is divided into two machines flow shop, flexible flow shop and re-entrant flow shop. The optimization models and the scheduling algorithms of the three production environments are studied respectively. Then, considering the subcontracting jobs and the accuracy of arrival information, this paper proposes a hierarchical production scheduling mechanism and constructive corresponding scheduling algorithms for each level respectively. The research contents of the four modules as flows:(1) For the flow shop problem with two batch-processing machines, three mixed integer linear programming (MILP) models are proposed to minimize makespan, maximum lateness and total tardiness, respectively. Three model-based heuristic algorithms are established to improve the computational efficiency of the corresponding models, respectively. Small-size instances are designed to demonstrate the optimal solutions can be obtained by the MILP models and the heuristic algorithms, respectively. Large scale instances are randomly generated, to demonstrate the computational efficiencies of the heuristic algorithms are higher than the direct solving the MILP models.(2) For the flow shop problem with parallel batch-processing machines and re-entrant jobs, a MILP model is developed to minimize makespan. Small-size instances are designed to test the performance of the MILP model. Two improved heuristics and a new constructive heuristic are proposed for large scale problem. The time complexities of these three heuristics and the worst-case of the new constructive heuristic are analyzed in detail. Compared with two improved heuristics by large scale instances, the performance of the new constructive heuristic is superior.(3) For the flexible flow shop problem with incompatible job families, a mixed integer programming model is conducted to minimize total weighted tardiness. Two improved heuristics and a new constructive heuristic are proposed to find approximate solution. The time complexities of these three heuristics are analyzed in detail. In order to test the efficiency of the heuristics, another typical heuristic algorithm is proposed. An extensive simulation study is conducted. The results show that the performance of the new constructive heuristic is better, which can meet the needs of practical application of the enterprise.(4) Consider the incompatible job families, the subcontracting jobs and the dynamic accuracy of arrival information in the flexible flow shop. A hierarchical production plan structure is proposed to minimize manufacturing cost. Three decision making levels include balancing capacity and demand, machining at the quenching stage and machining at the quenching stage. Three heuristics are established for the three corresponding decision making levels. An extensive simulation study is conducted to verify the effectiveness of the proposed strategies. The results are promising as compared to benchmark control strategies.