CLSP问题的分枝定价算法

提出了一种新的算法 分枝定价(Branch and Price)算法解经典CLSP,带有能力约束的单级多项动态批量问题(Thecapacitatedsingle level,multi item,dynamiclot sizingproblem)·CLSP问题有广泛工业背景,而且已被证明为NP Hard问题,它的目标是最小化总的装设(set up)费用和库存费用之和在所考虑的时间范围(horizon)内,并且满足给定约束条件·分枝定价算法是一种广义分枝定界(branch and bound)算法,它允许应用列生成(columngeneration)过程于整个分枝定界树·详细描述了该算法的实现,...

Branch-and-Price Algorithm for CLSP

A Branch-and-Price approach was presented to solve the classical capacitated single-level multi-item dynamic lot-sizing problems, also called as CLSP. Its objective is to minimize the sum of set-up and inventory-holding costs over the horizon under consideration, such as demands, capacity restrictions, etc. CLSP is frequently encountered in the most industry settings and well known as NP-Hard problem. The Branch-and-Price approach, a generalization of branch-and-bound with LP relaxation, allows column generation to be applied throughout explored nodes in the branch-and-bound tree. An implementation of the branch-and-price approach was described in detail. Two groups of examples were computed to verify the correctness and advantages of the suggested approach.