Multi-objective Hybrid Metaheuristic For Solving Hybrid Flow Shop Scheduling Problem

The level of production scheduling is a important factor of the modern manufacturingproduction process can be stable and efficient operation. Based the technical review on thedomestic and foreign research of shop scheduling and multi-objective optimization problems,this paper propose using non dominated sorting genetic algorithm with elitist strategy(NSGA2) to optimize multi-objective job shop scheduling problem.Combined with the actual environment of production workshop, it built amulti-objective flexible flow shop scheduling model including the makespan and the totaltardiness. When we use non-dominated sorting genetic algorithm with elitist strategy (NSGA2)to solve multi-objective optimization problems, found that computational efficiency andconvergence speed of the NSGA2algorithm is not satisfactory, and the parameter setting isvery complicated, there will be premature phenomenon. In order to solve this problem, thisarticle from two aspects to NSGA2algorithm is improved, and put forward the FLC-NSGA2algorithm and L-NSGA algorithm. The FLC-NSGA2algorithm, which makes use of fuzzylogic controller to change the crossover and mutation probability. The L-NSGA algorithm,which uses Lorenz dominance relations replace the Pareto dominance relations of NSGA2,since the Lorenz optimal front is a subset of the Pareto optimal front, so it can reduce thenon-dominated front size of NSGA2algorithm.This paper used the three algorithms to solve the multi-objective flexible flow shopscheduling, and analysis, comparison of the results, which show the FLC-NSGA2algorithmand L-NSGA algorithm has better search ability and higher efficiency than NSGA2algorithm, it certain guiding function to production practices.