考虑库存信息的钢管批量计划优化算法

摘要

A steel tube batch plan problem based on inventory information is studied .For the problem , a batch planning model is established with the objective of minimizing the billet shuffles .Combined with the feature of the problem , a two-stage optimization algorithm based on the billet continuous shuffling is de-signed:first, determining the current optimal billet and searching the billets in the lower floor on the same stock to match the best continuous rolling billet;second, improving the first stage solution for searching the continuous rolling billets in each rolling location .Through experimental verification based on the actual production data , the billet shuffles decrease significantly relative to the classic heuristic algorithm .The model and algorithm are proved feasible and effective .%在考虑库存信息的基础上对钢管批量计划优化问题进行研究，建立了批量计划轧制位置与垛位钢坯多对多关系的批量计划模型，实现对倒垛次数的优化。结合问题特征，设计了一种基于钢坯连续倒垛的两阶段批量计划优化算法：第1阶段，确定当前最优钢坯，进而搜索同垛位上的下层钢坯，匹配最佳的连续轧制钢坯；第2阶段，针对每个轧制位置搜索垛位上的连续轧制钢坯，改进第1阶段的解。通过基于实际生产数据的实验验证，相对经典启发式算法，倒垛次数显著降低，算法和模型是可行且有效的。