A Study For FFSP With Variable Processing Time

Flexible Flow shop Scheduling Problem is a classic problem in the production scheduling area.It has both features of classic flow shop and parallel machine shop.All jobs follows the same processing sequence at least one stage owns at least two identical machines which have same capacity.FFSP exists in automobile production,semiconductor packaging,steel production and large equipment manufacturing.Because of contains distribution of parallel machine and the processing sequence of jobs,FFSP is difficult to solve and it is a typical NP-hard problem.Recently,flow shop scheduling problem with variable processing time become a research hotpot in the fields of production scheduling.Different from the classic FFSP,the processing time in FFSP-VPT is a "flexible" parameter rather a "rigid" one.Processing time can be changed within a certain range of time based on user requirements and the situation of the indent.As the processing speed increases,the quality of the products will decrease,resulting in the increase of rework rate and the burden of production in the workshop.In the end,It delays the completion time.This paper mainly studies on the processing time under the condition of the rework rate,how to shorten the completion time,so as to better play to the role of the fine workshop production scheduling,workshop production efficiency,reduce production costs.Based on the research background of automobile production,this paper mainly studies flow shop scheduling problem with variable processing time.A mathematical programming model for FFSP-VPT is established,and the model parameters,notations and basic constraints are designed.At the encoding and decoding step,this paper proposes two-step encode mode based on the ROV,individuals are decoding as the processing sequence of jobs and the selected processing time number.First in first out and first available machine rules are used to arrange processing.In the global optimization Algorithm,this paper adopts the global optimization capability and the Bat Algorithm(BA)as the global optimization Algorithm.This paper adopts Bat Algorithm which has a strong optimization capability and high accuracy as the global optimization algorithm.In the study of bat algorithm,it is found that its evolutionary process is guided by a single individual,which can lead to a lack of population diversity and a local extreme value.As evolution progresses,BA's search range is constant,which can lead to bat deviation from optimal location to reduce search efficiency and reduce optimization accuracy.In order to solve the above problems,this paper improves the BA algorithm,and proposes a self-adaptive elite bat algorithm.SEBA proposes a best individual set which contains high fitness but low similarity of individuals based on hamming distance,an elite selected randomly from the set guide the evolution that can avoid the evolution step evolved into local extremum.Proposes an adaptive update mechanism,reduce the speed of location update along with the development of evolutionary namely narrow your search,the influence of the speed of the effect of location updating weights,made the bat better on a small scale to search the optimal solution.In order to make SEBA perform better,crosstest is designed to get the optimal value of the algorithm parameters.Aiming at the optimization problem of the flexible flow workshop in the process of variable work with converging production line,a solution based on inversion production algorithm and intelligent optimization algorithm is proposed.Firstly,the production mode of the main production line of the workpiece is determined by optimizing the algorithm,then the processing order of the auxiliary production line is solved by the method of inversion,and then the overall solution is obtainedThis paper mainly study flow shop scheduling problem with variable processing time.By analyzing and improving the shortcomings of BA,this paper propose a SEBA.The results of simulation results show that SEBA has greatly improved the evolution,precision and efficiency of solving problems.To solve FFSP-VPT with converging production line,this paper put forward based on the inversion algorithm of producing SEBA successfully solved this problem,and the actual problems in the automobile production workshop overall solution is given.