Storage Assignment For Multi-floor Circulated Distribution Center Under The Influence Of Elevator Queueing System

It is not a simple superposition of several single distribution centers.Due to the influence of vertical transport,the sorting operation of products in multi-floor distribution center is more complex than that in single distribution center.Hence,a storage assignment optimization model for the multi-floor distribution center is presented based on the idea of modeling in stages and in the context of considering of elevator's influence.At present,a few of imperfections about elevator queuing system and multi-floor assignment problem are found,such as "the strong assumption of elevator queuing system is not practical,but it is difficult to derive performance index under weaker hypothesis","the multi-floor problem does not consider the effect of elevators in the down-pattern",etc.Hence,in the first place,an elevator queuing model closed to actual is developed based on the Phase-Type distribution,a distribution with good analyticity and popularity.The queuing system of elevators in the down-pattern is described as a series of batch-arrival and batch-service queuing model,it can be expressed as PH(?)/PH(0,?]/1/C,C =?+C1.And it is equivalent to PH(x)/PH/1/C queuing model,whose arrival interval distribution is related to the system state,x.Then the analytic expression of average elevator waiting time and queue length at every floor are derived in order to quantitatively express the impact of the elevator with the help of the QBD process and matrix geometric method(MGM).Secondly,a three-stage with feedback loop storage assignment optimization model for the multi-floor distribution center is presented in the context of considering the elevator's influence.The products are divided into several product groups according to their own characteristics.Then each floor in the multi-layer distribution center is assigned a product group.Finally,the products within each floor are assigned to a certain location.There are different objectives in these three stages.The first is maximizing the product correlation in each group,the second is minimizing the products' centre of gravity and the last one is making the total picking time minimum.In addition,the algorithm is designed for each model such as genetic algorithm and differential evolution algorithm and hybrid differential evolution algorithm.Finally,multi-floor distribution center A is taken as an example to verify the validity of the elevator queuing model,the storage location assignment model and their algorithms.According to the results of model analysis,many factors should be considered in the distribution center when goods assigned.For example,the heavier products and faster moving products could been stored in lower floors to reduce the gravity and transportation distance,so as to improve the efficiency.In addition,it should be taken in consideration when determining the number or load of elevatoes that the total sorting time is minimum,rather than the elevator waiting time of one floor.