| With the characteristics of diversified products,special working mode and complicated working flow,the auto harness warehouse,as a special type of production-orient warehouse,is widely used in automobile parts enterprises.Auto harness warehouse is a complex and organic automation system,the overall system efficiency directly depends on the following two core elements: the policy of storage location assignment and the method of sequencing.So far,little attention was paid to auto harness warehouse,and therefore only a small amount of relevant research outputs were published.Thus,it is of great significant to do research on the storage location assignment and sequencing problem in the auto harness warehouse relevant for both theoretical development and practitioners' needs.The background,research status and development trend of storage assignment and sequencing problem in automated warehouse are introduced in this paper.Based on the layout and special working mode of the auto harness warehouse,mathematical models are constructed and several intelligent optimization algorithms have been developed.Computational tests on a real auto harness warehouse and a set of generated instances are applied to compare the overall performance under different parameters.The main research contents are as follows:Firstly,the overall arrangement and working mode of the auto harness warehouse are introduced in detail and the chief weaknesses are listed after analyzing the current status of manufacture and operation in the warehouse.Crucial and practical information hidden in the historical production data of the auto harness warehouse is explored,which lays a solid foundation for constructing the model of storage location assignment problem.A mathematical model which seeks out the optimal allocation of items is proposed so as to minimize equivalent travel time required to fulfill the production orders by a storage and retrieval machine.In order to solve the contradiction between local optimum and convergence rate of genetic algorithm,an improved genetic algorithm is developed for the proposed model.At the same time,a heuristic algorithm is proposed to determine the location for the incoming unit load.The behavior of proposed storage allocation algorithms is investigated under different scales of instances and different related parameters,and the results are compared with full turnover-based storage policy,random storage policy and class-based storage policy.This paper presents a discrete event simulation model that reproduces the behavior of the auto harness warehouse based on the result of storage location assignment and the operation strategy of the storage and retrieval machine.A hybrid simulated annealing and discrete particle swarm optimization algorithm is developed after redefining the position and velocity of particle.The proposed cooperative evolutionary algorithm is characterized by its fast convergence rate and global convergence.And the real data obtained from the auto harness warehouse in Delphi which triggered this research has been studied thoroughly.In the end,summary is given and some questions which need more attention in the future are raised. |
