基于钢铁企业的合同计划管理要求,建立了以产能平衡和最小化拖期提前总惩罚为目标的多目标数学规划模型,综合考虑了工序的前序关系、工序的产能和库存的约束。将生产合同与生产工序转换为节点图,从而将合同计划问题抽象成一种改进的旅行商问题。根据模型和问题的特点设计了带交货期启发信息的蚁群算法,并以钢厂实际合同数据为例进行实验。结果表明,改进蚁群算法获得的最好解和计算成本都比较令人满意,模型和算法是有效的。
Based on the management goals of order planning of the iron-steel plants, a mixed integer programming model for the order planning is constructed. The goals include balancing production capacity and minimizing the total cost of infringing contracts. The constraints include relationship between different processes, production capacity of machines and inventory level. As the data of orders and processes are put together to make a point graph, the problem of order planning is transferred into an improved Traveling Salesman Problem. An improved ant colony algorithm is presented to solve this model. The computational results show that the model comes up to the production processes, the satisfying solutions can be obtained within acceptable time.