外购板坯设计问题的优化模型及算法  

Optimization model and algorithm of the outsourcing slab design problem

作　　者：贾树晋[1] 杜斌[1] 黄可为[1] 李维刚[1] 

机构地区：[1]宝山钢铁股份有限公司研究院,上海201900

出　　处：《冶金自动化》2015年第2期22-27,40,共7页Metallurgical Industry Automation

摘　　要：针对外购板坯规格已知、合同质量需求为区间值、考虑工艺路线约束的板坯设计问题,建立了最小化总余材量的线性混合整数规划(MILP)模型,并提出了一种基于MILP的启发式算法,将大规模板坯设计问题分解成多个小规模MILP问题,使之易于优化求解。算法兼顾了启发式算法的快速性及MILP的最优性,能在较短的时间内得到近优解。现场数据测试表明该模型和算法具有良好的优化性能和实用性。As for the outsourcing slab design problem,in which the size of slabs is fixed,the order de-mands for weight are interval values,and the processing constraint is considered,a Mixed Integer Lin-ear Programming （ MILP） model to minimize the total surplus weight is built,and a MILP-based heu-ristic algorithm is also proposed. The large-scale slab design problem is decomposed into several small-scale MILP problems,which are easily solved,by means of a heuristic method. This algorithm takes advantage of the rapidity of heuristic method and the precision of MILP,and can obtain a near-optimum solution in a short time. Computational experiments with practical data demonstrate that this algorithm has good optimal performance and practicability.

关 键 词：板坯设计 线性混合整数规划 启发式算法 生产管理 

分 类 号：TF089[冶金工程—冶金物理化学]