A multi-objective operation optimization method for dynamic control of reservoir water level in evolving flood season environments

Current multi-objective optimization methods, traditionally rooted in static models, often neglect uncertainties and environmental interactions such as forecast accuracy and reservoir conditions. This study introduces a novel multi-objective operational optimization model aimed at dynamically controlling reservoir water levels in evolving flood season environments. The proposed model conducts a comprehensive analysis, quantification, and prediction of water level control dynamics during flood seasons by integrating strategies that encompass runoff forecast acquisition, dynamic risk assessment, and adaptive decision-making responses. To enhance the model’s effectiveness, this research proposes the Dynamic Multi-Objective Multi-Strategy Co-evolution (DMMC) algorithm. This algorithm incorporates several strategies, including memory-based individual optimal adaptation, dynamic updating of diverse individuals, collaborative updating based on forecast data, and static optimization techniques. These strategies enable real-time monitoring, identification, and efficient response to environmental fluctuations, thereby optimizing the sustainable utilization of water resources. Numerical experiments and engineering case studies validate the efficacy of the proposed method, demonstrating its capability to accurately capture environmental trends and promptly respond to evolving conditions. The simulations confirm the rationality and reliability of the model, presenting a novel approach for effectively managing dynamic water level control during flood seasons.