| Logistics centers are key departments and places of modern logistic activities. Order picking is considered as the pivot in logistics center, as it costs above50percent of total cost of the center. It is shown by researches and practices that the reduction of picking route travelled by pickers can be obtained by the efficient use of four strategies namely warehouse layout, storage assignment, order batching and picker routing. In this way picking time can be reduced and lead to the improvement of customer service level. This paper aims to improve work efficiency of picking task in manual picking system through designing reasonable strategy for order batching. Order batching is a process that combines several single customer orders into one batch or a larger order so as to increase the utilizations of equipments and reduce work load, so that picking process can be effectively carried out. Order batching problem is proved to be NP-hard problem, so study on the use of efficient heuristic algorithms is popular.This paper hopes to minimize the distance of picker routing via studying on order batching strategy. So that picking time can be saved and flow period of goods can be shorter, response to customer order can be faster. Upon previous studies the paper proposes three algorithms which called Reduce Batching Number RBN, PSO-based Batching Method PSOBM and RBN-based GA Batching Method RGABM separately. RBN improves batching process through considering the impact of batch number. PSOBM considers using particle swarm optimization method to solve batching problem, for matching the algorithm and the problem to be solved, the paper modifies binary particle swarm optimization. The proposing of RGABM benefits from previous works on the problem. Batch reducing process is combined with GABM to acquire better searching property. Genetic algorithm owns well convergence and optimizing ability, so among three algorithms it has the best performance. This paper also designs a way of calculating routing distance under S-shape strategy. At last, a simulation experiment upon matlab platform is carried out. Algorithms proposed are compared with classic heuristics. The result reveals that performances of three algorithms are better than the classic ones. Especially the group-based algorithms have remarkable improvement in optimizing property. |