Modified Genetic Algorithm For The Distributed Permutation Flow-shop Scheduling Problem

With the globalization of production and the scale of manufacturing enterprises,the distributed production scheduling is widely used in various large manufacturing enterprises and has become a key research issue in the field of job shop scheduling.The goal of the distributed job shop scheduling problem is to rationally allocate and sequence the jobs under various constraints to achieve the desired performance goals.However,there are few theoretical studies on distributed workshop scheduling problems,and integer linear programming models for distributed job shop scheduling problems need to be improved.Algorithms for solving distributed job shop scheduling problems are also very limited,and its are easy to fall into a local optimum prematurely.In order to minimize the maximum completion time of the distributed scheduling problem and solve the distributed production shop scheduling problem,this paper uses an improved genetic algorithm for the distributed job shop scheduling problem.Firstly,the K-means method is used to cluster the jobs according to the earliest completion principle,and uniformly distribute the clustered jobs to each job-shop.Then NEH and stochastic methods are used in combination to sort the jobs allocated to the job-shop and complete the initialization of the population.Secondly,Then,calculate the fitness and select the elite chromosomes,construct and update the probability matrix according to the selected elite chromosomes,mine the blocks according to the composition rules of the blocks to form a block library,and use the roulette according to the information in the block library and the probability matrix to construct artificial chromosomes and inject artificial chromosomes into artificial chromosomes.Thirdly the two methods are used to recombine genes based on plant-level chromosome fragment crossover and plant-based workpiece order-based chromosome fragment crossover to improve the diversity of solutions.To verify the optimality of the algorithm,the experimental results are compared with other well-known evolutionary algorithms to prove the effectiveness of the proposed algorithm.