随机型联合补充问题的混合遗传算法的研究

The Research of Hybrid Genetic Algorithm for Stochastic Joint Replenishment Problem

随机型联合补充问题是一种需求为随机的多种物品联合补充,目标是得到最小的总成本.针对这一问题使用了定周期检查策略进行分析,并建立了期望值模型,求出使整个库存系统的单位时间总期望费用最小的基本订购周期、各物品订购周期和最大库存量,并要满足一定的服务水平.对此模型采用了随机模拟算法和遗传算法的混合来进行解决,该算法可以得出问题的近似最优解,并用实例对算法进行了验证.

Stochastic joint replenishment problem is the joint order of a group of items with stochastic demand, our purpose is to acquire the least total cost. We analyze the problem using periodic review policy, and present an expected value model, acquiring basic order period, order period and the largest inventory of each item, and under the service level constraints. We propose a hybrid genetic algorithm which combines with a stochastic simulations algorithmic and a genetic algorithmic for solving this model, and it can acquire a near optimal replenishment policy. A numerical example is presented to illustrate the effectiveness of algorithm.