Automated Large-scale CVRP Solver Design via LLM-assisted Flexible MCTS

2026-05-05

AI Summary

This work addresses the large-scale capacitated vehicle routing problem (LSCVRP), which remains challenging due to its immense scale and the reliance of existing divide-and-conquer approaches on extensive handcrafted decomposition strategies and sub-solver configurations. To overcome these limitations, the authors propose LaF-MCTS, a novel framework that deeply integrates large language models (LLMs) with a flexible Monte Carlo tree search (MCTS) mechanism. LaF-MCTS features a three-level decision hierarchy that automatically designs both decomposition strategies and sub-solvers, enhanced by semantic pruning and branch regeneration techniques to efficiently navigate the algorithm design space. Experimental results on the CVRPLib benchmark demonstrate that the solvers automatically generated by LaF-MCTS outperform multiple state-of-the-art CVRP solvers, confirming the framework’s effectiveness and scalability.

Abstract

Solving large-scale CVRP (LSCVRP) with hundreds to thousands of nodes remains difficult for even state-of-the-art solvers. Divide-and-conquer can scale by decomposing the instance into size-reduced subproblems, but designing decomposition logic and configuring sub-solvers is highly expertise- and labor-intensive. Large Language Models (LLMs) have emerged as promising tools for automated algorithm design. However, existing LLM-driven approaches struggle with LSCVRP primarily due to the difficulty in generating sophisticated search strategies within a limited context window. To bridge this gap, we propose the LLM-assisted Flexible Monte Carlo Tree Search (LaF-MCTS), a novel framework that automates the design of high-performance LSCVRP solvers. We develop a three-tier decision hierarchy to enable incremental design of decomposition policies and sub-solvers for LSCVRP. To enable efficient search within the algorithmic hypothesis space, we introduce semantic pruning to eliminate semantically and structurally redundant codes, and branch regrowth to regenerate codes and preserve diversity. Extensive experiments on CVRPLib demonstrate that LaF-MCTS autonomously composes and optimizes decomposition-enhanced solvers that surpasses various state-of-the-art CVRP solvers.

Problem

• Large-scale CVRP
• Automated solver design
• Algorithm decomposition
• Vehicle Routing Problem
• Scalability

Innovation

• LLM-assisted algorithm design
• Flexible Monte Carlo Tree Search
• Large-scale CVRP
• Semantic pruning
• Branch regrowth