Hybrid heuristic for the production replanning problem under varying demands in manufacturing industries.

This study aims to solve a one-dimensional multi-period cutting stock production replanning problem encountered in manufacturing industry cutting departments, considering pattern-setup costs. For operational purposes, long-term coordinated production and cutting plans are often created based on predicted demand. However, real demand may differ from predictions, and replanning may be warranted by changing conditions. A mixed-integer mathematical model was proposed by minimizing both the production cost and the cost of deviating from established plans. In addition, a hybrid dynamic programming-based heuristic was proposed. Extensive experiments were conducted on randomly generated instances. The influences on the solutions are analysed by varying the holding cost per unit and the deviation cost. Finally, the proposed heuristic is applied to enterprise practice, and the results indicate that the total cost can be expected to be reduced by 14.55% on average. [ABSTRACT FROM AUTHOR]