钢铁企业产成品物流控制中转库流向的优化问题

Re-deposit Destination Optimization for Finished Product Logistic Control in a Steel Work

以我国大型钢铁企业上海宝山钢铁集团 (公司 )为背景 ,描述了其产成品物流控制中的产成品转库问题 ,建立了转库流向优化问题的数学规划模型·使用遗传算法 ,提出一种称为“两段式可重复非负整数编码”的编码策略来获得大规模 0 1规划模型的近优解·计算结果表明 ,当可用能力足够大并且准发数较少时 ,结果与SAS/OR的计算结果非常接近·当可用能力不是足够大 ,并且准发数较多时 ,计算结果要好于实际调度结果 ,此时SAS/OR软件在规定的时间内无法算出优化解·

The re deposit operation in Bao Steel,a large steel complex in China, was described and a mathematical programming model was built for the finished product re deposit problem. For this large scale 0 1 programming model, available software packages can't give out optimal solution in given time. A 'two stage, repeatable non negative integer coding' genetic algorithm was brought up to get the near optimal solution. When there are enough delivery capacity and the number of the sub order is small, the instance testing results of GA of is like that of the SAS/OR software; whereas when there are no enough delivery capacity and number of sub order is very large, the results of GA are better than real time scheduling results,and the SAS/OR software can't provide optimal solutions in given time.