Intelligent Methods Based Automated Warehouse Systems Optimization And Application

As a new storage and retrieve technique in the field of modern industry,automated storage and retrieval system(AS/RS) has the features of saving labor, highefficiency and utilization rate, which has already become an important part of FlexibleManufacturing System(FMS) and Computer Integrated Manufacturing System(CIMS).Moreover, AS/RS is an interdisciplinary forefront subject, involving many fieldssuch as computer science, automation, mechanism, electronic science, communicationtechnique, intelligent optimization theory and so on. So it is of great theoretic andpractical significance to investigate AS/RS.This paper addresses the location optimization, the warehousing schedulingsingle-objective optimization, the warehousing scheduling multi-objectiveoptimization. Meanwhile, performance evaluations of AS/RS with multi-asile andmulti-shuttle is also explored. As AS/RS is discrete-time system, and it belongs to akind of the combinatorial optimization problem and also is NP-Hard. At the sametime, the complexity of conventional precise algorithm has exponent relation to thesize of the problem. Therefore, it is difficult to obtain the optimal or suboptimalsolution within an acceptable period of time. Given this, The relativeoptimization-oriented problems are studied by using the intelligent optimizationtheory, on the basis of which the software platform is built to validate the feasibilityand effectiveness of the proposed methods. The main contents are summarized asfollows:Firstly, to overcome the drawbacks of storage location optimization only on thebasis of quantity of orders. A mathematical model is built by means of mathematicalstatistics, which represents the average run-time of a stacker within a production cycle.To obtain the optimal solution, a new bacterial foraging algorithm is proposed. Theelimination and dispersal is designed according to the level of contributing todiversity of population, which increases solution space and overcoms the drawback oftrapping in local optimal. Chemotactic stepsize is adjusted adaptively, which enhancesthe convergence efficiency. Furthermore, the population scales influencing the performance of algorithm are investigated and the convergence is proved. Theperformance of proposed algorithm is tested through simulation combined with theindustrial real world case, and results show that the algorithm is very efficient.Secondly, on a single objective of warehousing scheduling, a new shuffled frogleaping algorithm with crossover operator and heuristic mutation operator is proposed.Crossover operator which prevents evolutionary stagnation effectively. Heuristicmutation operator improves global search ability and leads sub-populations to searchin promising direction. Thus convergence speed is improved. The simulation andperformance analysis demonstrate the feasibility of this algorithm.Thirdly, aiming at the demands of production efficiency and quality control of anenterprise, a multi-objective constrained optimization model is founded. Due to theconflict among objectives, an improved multi-objective tabu search algorithm basedon pareto is proposed. A technique based on matrix theory is designed which is usedto produce neighborhood and greatly increase the diversity of neighborhood. Thepenalty strategy is proposed to reduce the the risk of falling into local optimalm. Thesimulation results show that the solutions produced by the proposed algorithm aredominant relative to other algorithms and their distributions are also uniform.Therefore, it proves the efficiency of the proposed algorithm.Fourthly, in order to evaluate the performance of warehouse with multi-asile andmulti-shuttle and make the relations among them clear. The efficiency of stacker isinvestigated under the variable number of asile, shuttle and velocity, and the valuationmodel is proposed. In the meantime, the efficiency of stacker is also studied underwarehouse of same volumes but different size. The conclusion shows that the moreefficient the stacker the closer to square the size of warehouse. It have an instructionrole for designing warehouse and arranging production plan.Finally, the testing platform of AS/RS is constructed based on C/S structure anddesign principle of modularization. Combined with the field in the enterprise, theperformance of AS/RS is tested and validated. It not only ensures the suitability ofAS/RS, but also provides references for the feasibility of industrial applications.