arXivDaily arXiv每日学术速递 周一至周五更新
arXiv周末暂无论文更新,休息一下吧,周末愉快~~
arXiv 2601.06502cs.AI
收藏

DRAGON:基于大语言模型的分解与重建代理用于大规模组合优化

DRAGON: LLM-Driven Decomposition and Reconstruction Agents for Large-Scale Combinatorial Optimization

Institute for Infocomm Research A STAR Singapore;Singapore Management University Singapore;Nanyang Technological University Singapore;Eindhoven University of Technology Eindhoven Netherlands;National University of Singapore \& Institute for Infocomm Research A STAR Singapore;Singapore University of Technology;Institute for Infocomm Research;Singapore Management University;Nanyang Technological University;Eindhoven University of Technology;National University of Singapore \& Institute for Infocomm Research

Shengkai Chen, Zhiguang Cao, Jianan Zhou, Yaoxin Wu, Senthilnath Jayavelu, Zhuoyi Lin, Xiaoli Li, Shili Xiang

AI总结:

DRAGON是一种结合元启发式设计和LLM推理的新型框架,通过分解与重建代理实现大规模组合优化问题的高效求解。

摘要:

大型语言模型(LLMs)最近通过基于提示的策略展示了在解决组合优化问题(COPs)中的潜力。然而,它们的可扩展性和泛化能力仍然有限,随着问题规模的增加,特别是涉及超过30个节点的路由问题,其效果会下降。我们提出了DRAGON,即分解与重建代理引导优化,这是一种结合元启发式设计和LLM推理优势的新框架。从初始的全局解开始,DRAGON能够自主识别具有高优化潜力的区域,并战略性地将大规模COP分解为可管理的子问题。每个子问题随后被重新表述为一个简洁、局部化的优化任务,并通过有针对性的LLM提示进行解决,这些提示由积累的经验引导。最后,局部优化的解决方案被系统地重新整合到原始全局上下文中,以产生显著改善的整体结果。通过不断与优化环境互动并利用自适应经验记忆,代理可以迭代地从反馈中学习,有效地将符号推理与启发式搜索结合起来。实验结果表明,与现有基于LLM的求解器局限于小规模实例不同,DRAGON在TSPLIB、CVRPLIB和Weibull-5k装箱基准上持续产生可行解,并在超过300万个变量的背包问题上达到近最优结果(0.16%的差距)。这项工作展示了反馈驱动语言代理作为通用和可解释大规模优化新范式潜力。

英文摘要:

Large Language Models (LLMs) have recently shown promise in addressing combinatorial optimization problems (COPs) through prompt-based strategies. However, their scalability and generalization remain limited, and their effectiveness diminishes as problem size increases, particularly in routing problems involving more than 30 nodes. We propose DRAGON, which stands for Decomposition and Reconstruction Agents Guided OptimizatioN, a novel framework that combines the strengths of metaheuristic design and LLM reasoning. Starting from an initial global solution, DRAGON autonomously identifies regions with high optimization potential and strategically decompose large-scale COPs into manageable subproblems. Each subproblem is then reformulated as a concise, localized optimization task and solved through targeted LLM prompting guided by accumulated experiences. Finally, the locally optimized solutions are systematically reintegrated into the original global context to yield a significantly improved overall outcome. By continuously interacting with the optimization environment and leveraging an adaptive experience memory, the agents iteratively learn from feedback, effectively coupling symbolic reasoning with heuristic search. Empirical results show that, unlike existing LLM-based solvers limited to small-scale instances, DRAGON consistently produces feasible solutions on TSPLIB, CVRPLIB, and Weibull-5k bin packing benchmarks, and achieves near-optimal results (0.16% gap) on knapsack problems with over 3M variables. This work shows the potential of feedback-driven language agents as a new paradigm for generalizable and interpretable large-scale optimization.

机构、总结、中文摘要均由AI分析而来,可能存在错误,请以论文正文为准。

↑