🤖 AI Summary
Existing automated heuristic design methods struggle to effectively accumulate and reuse search experience and rely on fixed evolutionary operators, lacking dynamic adaptability. This work proposes RefineEvo, a novel framework that introduces a planner-driven dynamic operator scheduling mechanism and a reflector-constructed bidirectional experience pool encompassing both positive and negative experiences. This enables state-aware, trajectory-driven adaptive heuristic evolution. By integrating large language models, evolutionary algorithms, and experience distillation techniques, RefineEvo significantly outperforms strong baselines across multiple classical combinatorial optimization benchmarks, achieving superior solution quality while substantially improving token efficiency. The framework advances heuristic design toward an experience-driven paradigm.
📝 Abstract
Automatic Heuristic Design (AHD) has emerged as a transformative approach for solving combinatorial optimization problems. While recent Large Language Model (LLM)-based methods have shown promise, they predominantly rely on fixed evolutionary operators and struggle to effectively accumulate and reuse historical search experience. This paper proposes RefineEvo, a novel evolutionary framework that transforms AHD from a static trial-and-error process into a planning-guided, experience-driven system. RefineEvo introduces a Planner to dynamically schedule evolutionary operators and trigger refinement based on the current search state, and a Reflector to distill valuable lessons into a Bidirectional Experience Pool containing both positive insights and negative pitfalls. This synergistic framework enables the system to adapt its search tools to the evolving complexity of the problem and leverage trajectory-aware, situation-conditioned insights to guide generation. Experiments on several classic combinatorial optimization benchmarks demonstrate that RefineEvo consistently outperforms strong baselines. In particular, RefineEvo delivers superior solution quality while improving token efficiency, enabling more efficient and autonomous heuristic design.