Research On Flexible Job Shop Scheduling Problem Based On AGA-DE

Shop scheduling problem is the most critical part of manufacturing execution system,and it is a research hotspot under the development environment of manufacturing industry.Good shop scheduling can optimize the production resources of the enterprise to the greatest extent and ensure the smooth operation of the shop effectively.Flexible Job shop Scheduling Problem(FJSP)is an extension of traditional Job shop Scheduling Problem(JSP).FJSP is more in line with the actual production shop than JSP.At present,among many algorithms to solve FJSP problems,genetic algorithm is the most widely used,but the traditional genetic algorithm has some defects,and differential evolution algorithm can make up for these deficiencies,and give full play to their respective advantages.Therefore,this paper proposes a kind of Adaptive Genetic Algorithm Differential Evolution(AGA-DE for short)to solve the FJSP problem.The main research contents are as follows:First of all,in the process of population initialization,its performance directly affects the efficiency of the algorithm and the overall performance of the algorithm.Therefore,in order to solve the problem of low population quality caused by random initialization,the traditional population initialization method is improved in this paper.After the initial population is generated randomly,the roulette strategy is used to screen the population,which can not only ensure the randomness and diversity of the initial population,but also improve the quality of the initial population and accelerate the convergence rate of the algorithm.Secondly,the genetic algorithm is more inclined to obtain the global optimal solution through crossover,while the differential evolution algorithm is more inclined to improve the local search ability through variation.Moreover,the crossover and mutation probabilities of most genes are fixed values,and the selection of these values is difficult,which is easy to cause the local optimization of the algorithm and reduce the convergence rate.Therefore,an improved adaptive crossover and mutation probability formula is proposed in this paper,so that the values of crossover and mutation probability will show corresponding linear changes due to the changes of individual and population fitness values.On this basis,the crossover and mutation probability values are adjusted,so as to improve the crossover and mutation modes in the process of evolution,so as to improve the global search ability of the algorithm,but also accelerate the convergence speed of the algorithm.Thirdly,in order to solve the problems that the genetic algorithm is easy to enter the local optimal and precocious in the late evolution process,a local convergence criterion based on real number coding is adopted in this paper,and the local optimal solution is quickly found and called out through adaptive differential evolution search.This index can track the convergence trend of the algorithm in real time at the late stage of evolution and judge whether the algorithm converges prematurely.If the algorithm reaches the local optimum,crossover and mutation of the genetic algorithm have almost no effect.In order to jump out of the local area search in time,the differential evolution algorithm with differential mutation operator should be implemented to generate new individuals.The value of the scaling factor can be changed adaptively according to the evolution of the population.The adaptive scaling factor can greatly reduce the complexity of the solving process,improve the solving efficiency,and accelerate the convergence rate of the algorithm.Finally,on the basis of using the test function to test the effectiveness of the improved algorithm,the improved algorithm is applied to the classical example of flexible job shop scheduling problem,and the simulation results of different algorithms are compared and analyzed,demonstrating the advantages of the method in the maximum completion time and convergence rate,and verifying the feasibility and effectiveness of the improved algorithm.Finally,the improved algorithm is applied to a practical case to prove its practical significance.