Title: Attraction-repulsion based balance of exploration and exploitation in many-objective optimisation with application to water resources allocation

Authors: Yingnan Ma; Junlong Shen; Di Zhu; Renbin Xiao

Addresses: School of Artificial Intelligence and Automation, Huazhong University of Science and Technology, Wuhan, Hubei, 430074, China ' School of Artificial Intelligence and Automation, Huazhong University of Science and Technology, Wuhan, Hubei, 430074, China ' School of Artificial Intelligence and Automation, Huazhong University of Science and Technology, Wuhan, Hubei, 430074, China ' School of Artificial Intelligence and Automation, Huazhong University of Science and Technology, Wuhan, Hubei, 430074, China

Abstract: This paper addresses the critical issue of water resource allocation in the Yellow River Basin, a region characterised by severe water scarcity and heterogeneous spatial distribution. A novel many-objective evolutionary algorithm, MaOEA/AROA, is proposed, integrating attraction-repulsion mechanisms and co-evolution strategies to effectively balance exploration and exploitation in optimisation. From the environmental, social, and governance (ESG) perspective, the model aims to minimise pollution, maximise equity, and enhance economic benefits, addressing the industrial, agricultural, and domestic water demands across nine provinces. The algorithm extends the single-objective attraction-repulsion optimisation paradigm to a many-objective framework, leveraging its strong exploitation capability while maintaining diversity through a dual-population co-evolution approach. Experimental results demonstrate that MaOEA/AROA generates a diverse set of Pareto optimal solutions, offering flexible strategies to balance conflicting objectives and promote sustainable development. The algorithm's performance is validated on standard benchmark problems and applied to the Yellow River Basin, showcasing its practical utility in complex water resource allocation scenarios. Future work will focus on enhancing the model's robustness and incorporating dynamic water cycle considerations to further improve its applicability in real-world water resource management.

Keywords: many-objective optimisation; attraction-repulsion optimisation; indicator-based MaOEA; co-evolution; water resource allocation.

DOI: 10.1504/IJBIC.2026.151784

International Journal of Bio-Inspired Computation, 2026 Vol.27 No.1, pp.31 - 44

Received: 10 Feb 2025
Accepted: 04 May 2025
Published online: 19 Feb 2026 *

Full-text access for editors Full-text access for subscribers Purchase this article Comment on this article