Research On Flexible Job Shop Scheduling Problem With Transport Time Consideration Based On Hybrid Genetic Algorithm

Manufacturing is an important pillar in our national economy.Under the background of "Industrial 4.0" and the diversified demand,the enterprise must make more reasonable scheduling scheme in order to respond to the market in time.The time other than the processing time is neglected in the study of the traditional workshop?scheduling problem,but the transport time between the machines exist in the actual production.Moreover,the transport time will change dynamically with the flexibility of the machine selection in the flexible job shop scheduling.Without considering the transport time,the optimization of shop scheduling has a great impact.Based on this,the paper takes the flexible job shop as the research object,and adds the transport time factor into the scheduling model to study the flexible job shop scheduling problem with transport time consideration.First,this paper analyzes the flexible job-shop scheduling problem with transport time consideration.Based on the traditional scheduling model,it makes new assumptions and establishes a new flexible job shop scheduling model with a minimum makespan,a minimum bottleneck load and a minimum load of the total machine for the optimization goal.Then,a hybrid genetic simulated annealing algorithm based on memory guidance is proposed to solve the new model.Different from the traditional serial hybrid structure,this paper uses the simulated annealing algorithm to realize the local optimization of the dynamic memory library,and then utilizes the optimized memory population as the guidance of the genetic algorithm,which blending the global and local searching ability of the hybrid algorithms.The genetic algorithm has been improved in the following aspects:For the sake of initial rules,the paper designs the minimum working time method considering the transport time is to improve the quality of the initial population.The crossover operator has a large probability to introduce the high quality population of the memory which improves the search speed of the algorithm.The mutation operator adjusts the mutation probability dynamically according to the change of the appropriate value,which prevents the algorithm from stopping prematurely.In the simulated annealing algorithm,it proposes a new neighborhood structure based on effective gene block to ensure the effectiveness of local search.For multiple optimization goals,this paper uses the improved Pareto order and the crowded distance of the niche technology to carry out the secondary distribution of turmoil which ensure that the algorithm converges to the uniform distribution of Pareto surface.Finally,the paper uses MATLAB to test the 13 common international cases with two kinds of transportation time,and takes the quality and distribution of the solution as the evaluation criterion of the algorithm and traditional genetic algorithm.Each group of examples is run 10 times in a row,received 520 groups experiments.Experiments show that compared with the traditional genetic algorithm,the hybrid algorithm always obtains the Pareto non-dominated solution with better quality and more uniform distribution.With the increase of the problem scale,the optimization of the hybrid algorithm is more obvious,which proves the proposed algorithm can effectively solve the problem of flexible job shop scheduling with transport time consideration.