Application Research Of Flexible Job-shop Scheduling Problem Based On Improved Genetic Algorithm

The development level of the manufacturing industry reflects the level of productivity of a country.The production workshop is the foundation of the manufacturing system.The optimization of production scheduling is the core of advanced manufacturing enterprises and modern manufacturing technology,and is the key technology to achieve high production efficiency and high reliability of enterprises.Effective workshop scheduling methods and optimization techniques have important theoretical and practical significance for the modernization of manufacturing enterprises.In this paper,various flexible job-shop scheduling problems(FJSP)are researched and explored,combined with genetic algorithm and particle swarm optimization algorithm to improve and integrate them.In this paper,three optimization algorithms are designed,and a prototype system for flexible job-shop scheduling problems is developed to provide theoretical guidance and technical support for the actual production workshop scheduling problems.For the singleobjective flexible job-shop scheduling problem,an improved variable neighborhood search hierarchical genetic algorithm is proposed,and the makespan is the goal.In the algorithm,the chromosome adopts a two-layer coding structure and uses a hybrid method to generate the initial population;divide the initial population into N subpopulations,perform improved genetic operations in each subpopulation,and store the obtained optimization results in the elite library.To prevent the loss of the optimal solution,adaptive variable neighborhood search is used in the elite library,and three different neighborhood structures are designed.During the iteration process,the neighbors with good optimization effects are adaptively selected for the next search,which promotes the competition between the neighbors,so that the neighborhood structure with better search results has a higher probability for algorithm optimization.For the multiobjective flexible job-shop scheduling problem,an improved genetic annealing algorithm is proposed to solve,and the three goals of makespan,critical machine workload and total machine workload are weighted to convert the multiobjctive problem into a singleobjective problem.In the algorithm,the crossover process uses an improved multiparental operation crossover method,multiparental generates multichildren,which realizes the recombination of genes and accelerates the convergence rate of the algorithm.Update the optimal individual library in time during the crossover and mutation process.Simulated annealing operation is performed on the optimal individual library after mutation,and a refined search is performed through the annealing mechanism to avoid the genetic algorithm from falling into the local optimum.Make full use of the advantages of genetic algorithm and simulated annealing algorithm,enhance the local search ability of genetic algorithm,and improve the efficiency of the algorithm.For the multiobjective flexible job-shop scheduling problem,a discrete particle swarm optimization algorithm with adaptive inertia weights is proposed,and the solved targets are Pareto optimal solutions of three goals: makespan,critical machine workload and total machine workload.During the evolution process,the discrete particle swarm optimization algorithm is used to directly solve the value of the next generation chromosome in the discrete domain.The location update uses the crossover and mutation operations in genetic algorithms.An adaptive inertia weight method is proposed,which adjusts the inertia weight according to the distance between the current position of the particle and the global optimal position,ensuring the balance of the global search and local search capabilities of the algorithm.Developed a prototype system for flexible job-shop scheduling problem,which was used to study and simulate the proposed optimization algorithms.The system can run simulation experiments on actual workshop problems and five sets of international standard calculation examples(5 Kacem problems,10 BRdata problems,21 BCdata problems,18 DPdata problems and 66 HUdata problems),and test and analyze the obtained simulation results to verify the performance of the three optimization algorithms in the paper.Finally,the research contents and innovations of this thesis are summarized,and the future research direction is prospected.