Research On Distributed Assembly Permutation Flowshop Scheduling Problem Based On Iterated Greedy Algorithm And Evolutionary Algorithm

As global economic integration and market competition continue to intensify,the unified management model of multi-site production is gradually applied in manufacturing.This production model brings a new scheduling problem — the distributed scheduling problem.This paper proposes intelligent optimal scheduling algorithms based on the iterated greedy and evolutionary algorithms for the distributed assembly permutation flowshop scheduling problem(DAPFSP).The specific work is as follows:Firstly,the background and significance of this paper are explained,and the current status of domestic and international research on the DAPFSP is reviewed and analyzed.Secondly,a mixed integer programming model is developed and a group-based iterated greedy algorithm is proposed for the DAPFSP with the optimization objective of minimizing the total flowtime.In the proposed algorithm,a heuristic algorithm is designed to obtain the initial population,destruction,construction and local search operators for products and jobs are designed to improve the search efficiency,and a hybrid selection method based on objective values and individual ages is used to increase the diversity of the population.Thirdly,an enhanced memetic algorithm is proposed for the DAPFSP with the optimization objective of minimizing the total tardiness.Based on the problem characteristics,a heuristic algorithm with a random operator is proposed to improve the initial population quality while maintaining the diversity of the initial population.A nested iterative structure and a population update method suitable for this structure are adopted to improve the efficiency of the algorithm.A new genetic operator and four local search operators are designed,and the adaptability of the operators is enhanced by adding pending parameters.Fourthly,a two-stage evolutionary algorithm is proposed for the DAPFSP with the optimization objective of minimizing both the total flowtime and total tardiness.A twopopulation structure based non-dominated sorting genetic algorithm is used in the first stage,and a heuristic algorithm with a two-population structure is proposed.Based on the characteristics of this structure,crossover operators,mutation operators and integration interaction methods are designed for different optimization objectives,which reduce the repetition within populations and increase the diversity among populations.In the second stage,a simplified decomposition-based multi-objective evolutionary algorithm is used to eliminate the tedious individual improvement stage,and an adaptive genetic operator is designed.Finally,the effectiveness and efficiency of the proposed algorithm and its key modules are verified by comparing with related scheduling algorithms on the authoritative literature.