渡线
数学优化
进化算法
调度(生产过程)
计算机科学
操作员(生物学)
帕累托原理
算法
遗传算法
数学
人工智能
生物化学
转录因子
基因
抑制因子
化学
作者
Binghai Zhou,Lingwei Zhao
标识
DOI:10.1016/j.cie.2022.108385
摘要
• Study the material-feeding mode based on the line-integrated supermarkets. • Focus on a two-level logistics network in mixed-model assembly lines. • Propose a MOEA/D-DFMB algorithm to solve the milk-run material-feeding scheduling problem. • Infuse several operators into the proposed algorithm to improve the solution quality. With the development of personalized and differentiated customization demands, mixed-model assembly lines have been increasingly adopted by manufacturing enterprises due to their low-cost and high-efficiency. However, this mode has brought great challenges to the material-feeding process which is highly coupled with the production of automobile assembly lines. This paper, therefore, proposes a milk-run material-feeding scheduling problem based on two-level logistics network for modern mixed-model assembly lines, focusing on the milk-run material-feeding process between the central warehouse and line-integrated supermarkets for the first time. An integer programming mathematical model is established with the objective of simultaneously minimizing the number of electric vehicles put into use and the maximum travelling distance of the electric vehicles. Then, a double-faced mirror boundary-based multi-objective decomposition evolutionary algorithm (MOEA/D-DFMB) is provided to solve the problem. The introduction of the double-faced mirror boundary theory helps overcome the uneven distribution of Pareto solutions. Besides, the adaptive crossover operator based on evolutionary potential judgment, the adaptive adjustment strategy of mutation probability, and the local search optimization operator are applied to the proposed algorithm, so as to enhance the search ability and the convergence speed. Finally, computational experiments of the proposed algorithm are carried out, and the feasibility and effectiveness of the proposed algorithm are evaluated by comparing with the non-dominated sorting genetic algorithm-II (NSGA-II), the strength Pareto evolutionary algorithm-II (SPEA-II), the multi-objective decomposition evolutionary algorithm based on differential evolution (MOEA/D-DE), the multi-objective red deer algorithm (MORD), and the multi-objective social engineering algorithm (MOSE). The results indicate that the proposed MOEA/D-DFMB outperforms the comparison algorithms both in solution quality and convergence rate when solving the milk-run material-feeding scheduling problem.
科研通智能强力驱动
Strongly Powered by AbleSci AI