基于学习机制的分布式混合流水车间调度算法研究

A Learning-Based Algorithm for Distributed Hybrid Flow Shop Scheduling

在现代制造业中,DHFSP对提高生产效率和降低成本具有重要意义。提出了一种基于迭代贪婪的自适应学习算法,旨在有效地最小化最大完工时间。首先,通过一种初始化策略构建初始解;其次,引入一个基于历史反馈的动态学习机制,该机制能够根据累积的优化经验,智能选择扰动策略,以逐步提升解的优化质量;最后,通过轮盘赌选择策略,进一步增强了算法对全局最优解的探索能力。实验结果显示,所提出的算法在求解分布式混合流水车间调度问题时,相较于现有的对比算法,展现出了显著的性能优势。

The distributed hybrid flow shop scheduling problem holds significant importance in modern manufacturing as it contributes to enhanced productivity and cost reduction.However,the intricate nature of the DHFSP poses a major challenge when seeking solutions.To address this issue,this paper proposes an adaptive learning algorithm based on iterative greedy that incorporates a learning mechanism with the objective of effectively minimizing the maximum completion time.The algorithm comprises three components:firstly,an initialization strategy is employed to construct the initial solution;secondly,the core of the algorithm lies in its dynamic learning mechanism that utilizes historical feedback to intelligently select perturbation strategies based on accumulated optimization experience,thereby progressively enhancing the quality of the solution;and lastly,a roulette selection strategy further enhances the algorithm's ability to explore the global optimal solution.The experimental results demonstrate that the algorithm proposed exhibits significant performance advantages over existing comparison algorithms in addressing the DHFSP.