Multi criteria decision-making method for optimal reservoir operation using multi-objective evolutionary algorithm and hybrid caramel algorithm

Abstract

Reservoir operation requires balancing competing objectives, including municipal water supply, irrigation, industrial demand, and environmental-flow requirements. This study develops an integrated decision-support framework that combines a decomposition-based multi-objective evolutionary algorithm (MOEA/D), the hybrid Caramel optimization algorithm, and an entropy method–compromise programming approach for optimal reservoir operation. The novelty of the framework lies in coupling evolutionary optimization with a structured multi-criteria decision-making procedure, enabling not only the generation of Pareto-optimal operating policies but also the systematic selection of implementable solutions that satisfy both water allocation and environmental objectives. The framework was applied to the Ravishankar Sagar Reservoir, India, using MOEA/D and Caramel algorithms to generate Pareto-optimal release policies. Model performance was evaluated using demand–satisfaction, deficit reduction, environmental-flow compliance, and reliability, vulnerability, and resilience indices. Results showed that both algorithms generated feasible operating policies; however, MOEA/D achieved higher demand satisfaction, lower annual deficits, improved reliability and resilience, and better environmental-flow compliance than Caramel. The compromise solution selected through multi-criteria decision-making achieved municipal, irrigation, and industrial demand–satisfaction rates of 93.5, 86.4, and 80.1%, respectively, for MOEA/D, compared with 89.7, 82.8, and 76.5% for Caramel. The proposed framework thus provides a transparent and transferable methodology for reservoir-operation planning.