Large-language-model-driven multiple algorithm evolution for network-constrained unit commitment model reduction

doi:10.1016/j.apenergy.2026.128745

Large-language-model-driven multiple algorithm evolution for network-constrained unit commitment model reduction

2026-09-01

0

PRE

AI

钟海旺 (Haiwang Zhong) *

C

Chongqing Kang

DOI：10.1016/j.apenergy.2026.128745

Original

Original request for help

Save

Abstract

En 中文

• An LLM-driven multiple algorithm evolution method is developed for the NCUC model reduction. • Code repair strategies and domain-specific knowledge embedding are incorporated to enhance the algorithm evolution method. • Existing NCUC model reduction algorithms are systematically combined with LLMs to improve generalization, accuracy and computational efficiency. • The infeasible solution repair algorithm is integrated to ensure solution feasibility after model reduction.

Large-language-model-driven multiple algorithm evolution for network-constrained unit commitment model reduction

Abstract

Journal

Organization