关键词: 自动化混流生产线, 并行制造单元, 调度, 能力约束, 重入工艺, 混合整数规划

Abstract: To improve the throughput of automated mixed production line,bottleneck machines usually have parallel manufacturing capabilities so that multiple parts can be processed simultaneously in a machine.Due to the fact that the actual processing time of parts on the parallel unit may span multiple cycles,an equivalent conversion method which transformed the object of using capacity constraints from the overall workstation to the independent units was proposed.On this basis,a Mixed Integer Linear Programming (MILP) model for cyclic scheduling problem in automated mixed production line was constructed,including the using capacity constraints of the parallel machine.Meanwhile,a mixed integer linear programming model for the problem without reentrance was constructed by borrowing the existing modeling approach for using capacity constraints on the parallel machine.Computational test on the improved benchmark instance demonstrated that the handling operations in the parallel unit had matched with each other alternatively as the cycle changes,and the feasibility of the proposed methodology was proved to formulate the workstation using capacity constraints.A large number of randomly generated instances verified the performance of the proposed model.Comparative experiments based on randomly generated instance demonstrated the optimality of the proposed methodology for using capacity constraints on the parallel machine.

Key words: automated mixed production line, parallel manufacturing unit, scheduling, capacity constraints, reentrant process, mixed integer programming