Suitability Research Between Order Structure And Order Picking System

Order picking which consumes the most labor within the distribution center, decides on the service level for downstream customer, and is the most important working process in distribution center. It plays a vital role to choose the right order picking system to improve the efficiency and improve customer service level. Selection of order picking system is of great significance, but at present, it is lack of necessary theoretical guidance and summary on the selection of order picking system. People rely mainly on their own experience and suggestion of suppliers to select proper picking system. At present there are some papers on control, scheduling, optimization, design about order picking systems, but it is still lack of research on suitability between order structure and order picking systems.The main reason for the lack of study on suitability is that it is complicated system engineering, and its complexity is embodied in four aspects. First is the complexity of the target:to choose a suitable order picking system, people not only need to take the efficiency of the system, one-time investment, daily operation expenses and other quantitative indexes into consider, but also have to consider the automation level, stability, comfort and so on qualitative indexes as the selection target. Second is the complexity of the constraints on choosing system:both order attributes and goods attributes on the constraints of the input factors, and the external conditions such as environmental attribute constraint should be considered, and these attributes themselves are very complicated. Third is the complexity of the order picking system:there are various order picking systems such as artificial equipment, semi-automation and automation equipment. And each type of specific equipment structure is very complex, even one vendor one structure. Fourth is the complexity of the strategy:in the selection process of order picking system, there is a consideration of many complicated strategies including layout design, storing, zoning, batching, routing and so on.Because of the complexity of suitability problem, this paper couldn’t answer all branches of these problems, but to choose the most crucial four difficulties as the research object. The first key point is to build up the system model to do the lateral comparison among many order picking systems. Second, comparison between different picking systems should be based on the comparison of respective optimal results. That is, at first, each model should be the optimal solution of the picking time, then to do the comparison before comparison. The different picking systems should choose the most appropriate pickying strategy; and the different order how to influence the choice of picking strategy also should be considered. The fourth is to do the research on relationship of order structure and picking system, the key is on the choice of picking system when the order intensity anf density begin to change.In the process of research on these four difficult problems, the main content of this article and the results are as follows:(1) Suitability problem between order structure and picking system is put forward, and some basic concepts of order matrix, order structure, order picking system, picking strategies, order matrix blocking and so on. According to the analysis of the logical relationship among pickers, orders and goods, all the picking systems can be summaried as three basic types that is picker-to-parts system, put system and parts-to-picker system. The time expression which is suitable to three general picking systems on one dimension environment is established. The space filling curves are used to link one-dimensional time expression blocks to form a two dimensional system model which can reflect the real world better. Three zoning strategies are summarized to put forward the corresponding time model figure. Set up a applicable problem model judged on picking time of total orders and sub-problem model which can optimize picking time.(2) The sub-problem of optimization about picking time is put forward. The two-dimensional blocking problem of order matrix is transformied into one dimensional coding, so that the batching and zoning optimization can be solved in the meantime using some classical heuristic algorithm such as Genetic Algorithm, Ant Colony Optimization Algorithm and Particle Swarm Optimization Algorithm. Comparing the computational time of three algorithm and the optimization effect, the genetic algorithm is finally chosen as the solving methods for this problem. According to preliminary optimization results, some features of order matrix blocking under different strategies are investigated. These features include vice diagonal characteristics for progressive zoning, total equilibrium characteristics for independent synchronized zoning and wave equilibrium characteristics for relevant synchronized zoning. A kind of quick distribution algorithm based on block characteristics is put forward to get the better initial solution in a very short period of time. The effect about the traditional artificial optimization, quick distribution algorithm and genetic algorithm is verified.(3) The zoning strategies and the suitability of picking system is discussed. The theorem that how the three kinds of system respectively choose three zoning strategies is proved using mathematical induction method. Some important conclusions are put forward, for example, "with the same given orders, the same zones number, the same batches number and the same optimization effect, the independent synchronised zoning is better than relevant synchronised zoning just considering the time" and "with specific conditions, considering the time, the synchronised zoning is better than the synchronised zoning". Through statistics on the simulation data, the theorems are verified, and the influence of order density, order intensity, zones quantity and batches quantity is analyzed.(4) The suitability between order structure and picking systems is discussed. The DEA model to solve the problem is constructed by considering the rational DMU (decision making units) and the input data and output data.Using this model, the choice of different picking systems with given conditions, the order density, order intensity, zones quantity and batches quantity which influence the selection of picking system is analysed. It is found that the order density, the zones quantity and batches quantity have significant influence on picking time, and the influence of order intensity is weak. Through the statistical analysis on simulation data, the order density is found the strong effect on picking system selection.This method not only has a great theoretical significance, but also has great application value. It can be able to guide the people in distribution center to select appropriate system according to order, and also can guide design and optimizing picking system which is adapt to the order, it has large application prospect in the industries of tobacco, pharmacy and tire.