湖北汽车工业学院学报2025，Vol.39Issue(3)：56-63,8.DOI:10.3969/j.issn.1008-5483.2025.03.010
基于Q-learning的改进NSGA-Ⅲ求解高维多目标柔性作业车间调度问题
Improved NSGA-Ⅲ Based on Q-learning for Many-objective Flexible Job Shop Scheduling Problem
摘要
Abstract
Aiming at the multi-variety and small-batch production mode in machining shops,a many-objective flexible job shop scheduling model was established to minimize total energy consumption,maximum completion time,machine load and total tardiness,and was solved by improved NSGA-Ⅲ.A triple encoding method based on machines,processes,and batches was adopted for encoding,and the initial population was generated using chaotic sequences produced by Logistic mapping.In addition,the reinforcement learning state space was constructed according to the quality index of the target solu-tion,and the neighborhood search strategy was adjusted through Q-learning training.Finally,simula-tion comparisons were conducted using benchmark instances and a manufacturing case,validating the effectiveness and superiority of the proposed model.关键词
柔性作业/目标优化/批量调度/Q-learning/邻域搜索Key words
flexible job/objective optimization/batch scheduling/Q-learning/neighborhood search分类
机械制造引用本文复制引用
张小培,陈勇,王宸,袁春辉..基于Q-learning的改进NSGA-Ⅲ求解高维多目标柔性作业车间调度问题[J].湖北汽车工业学院学报,2025,39(3):56-63,8.基金项目
国家自然科学基金(51475150) （51475150）
湖北省高等学校中青年科技创新团队计划项目(T20200018) （T20200018）
湖北省高等学校实验室研究项目(HBSY2024-80) （HBSY2024-80）
