Research On Uniform Parallel Batch Machines Scheduling Problem

As a kind of classical combinatorial optimization problem,production scheduling problem is widely exists in modern production industries for its important research value.Expanding from the classical scheduling problem,batch scheduling problem has become one of the hotspots in current research,because of its extensive practical value and good theory support.The rise of emerging production mode and its pervasive application in the enterprise makes the production environment more complicated than ever.The rapid development of information technology help the Internet of things technology is widely applied in the process of manufacturing enterprise,which brings opportunities but also brings challenges to the enterprises.As an important part of the new generation information technology,the Internet of things has great application value in the field of production scheduling.It can help to recognize the production scheduling objects,capture the state and location information of the production scheduling objects and machines,and so on.Make full use of the information to develop reliable and efficient scheduling policy and achieve intelligent optimization scheduling is pivotal for the enterprise to have an invincible position in furious marketing competitions.Based on the information provided by the Internet of things,this paper studied the uniform parallel batch machines scheduling problems under two special situations,which are abstracted form the processing procedures of semiconductor manufacturing industry.These two special situations include cases of different transportation time and machines with non-identical capacities.The scheduling objectives are both to minimize the makespan.The paper's main tasks are as follows.(1)The uniform parallel batch machines scheduling problem in consideration of the coordination of the production and transportation is studied.With the objective to minimize the makespan,we built the mathematical model of the studied problem.On the basis of analyzing the problem's related character,a heuristic algorithm and a local search strategy are put forward.Then,a hybrid DPSO-GA algorithm combining the advantages of a discrete particle swarm optimization algorithm(DPSO)and a genetic algorithm(GA)are proposed to solve the problem.In order to verify the effectiveness of the proposed DPSO-GA algorithm,we conduct the computational experiments.The experimental results show that the proposed hybrid DPSO-GA algorithm can effectively and efficiently solve the problem within a reasonable time.(2)The scheduling problem of minimizing makespan on uniform parallel batch machines with non-identical capacities is studied.A mixed integer programming model is formulated for the studied problem.In order to obtain a good initial solution,a heuristic is developed.Then,a genetic algorithm improved by a local search strategy is designed for the problem.Computational experiments are conducted and the results show that the improved genetic algorithm can effectively solve the proposed problem.