Study on Reservoir Ecological Scheduling Based on Multi-Objective Optimization

Hydropower development in river basins has significantly promoted economic growth while greatly changing the river ecosystems. Effective reservoir management is crucial to maintaining economic benefits while minimizing impacts on fish species. This study focuses on Reservoir X, which has annual regulation capacity, and proposes an ecological scheduling model for the fish spawning period using the NSGA-II algorithm combined with water temperature and TDG (Total Dissolved Gas) predictions. The model framework is as follows: first, hydrological analysis is conducted based on natural flow data at the dam site to determine the flow requirements for target fish species during their spawning period, providing constraints for optimization. Second, multiple regression methods are used to predict the discharge water temperature and TDG saturation of Station X. Finally, multi-objective optimization is performed considering hydropower generation, fish spawning period water temperature requirements, and TDG risks as objectives, with flow requirements during the spawning period, flood control, and water balance as constraints. The proposed model provides practical parameters for reservoir operation and guidance for different optimization objectives across various reservoirs.