Mathematical Modeling for Short-Term Hydropower Optimization: New Formulations & Predictive Approach скачать в хорошем качестве

Mathematical Modeling for Short-Term Hydropower Optimization: New Formulations & Predictive Approach

Séminaire “Une chercheuse du GERAD vous parle!” Mathematical Modeling for Short-Term Hydropower Optimization: New Formulations and a Predictive Approach Sara Séguin, Université du Québec à Chicoutimi, Canada 29 jan. 2025 Hydropower scheduling involves optimizing the operations of a hydropower system, which typically includes multiple turbines across interconnected power plants. Managing such a system requires solving a variety of optimization models, particularly the short-term hydropower scheduling problem. The solution to this problem determines the water discharges at each power plant, reservoir volumes, and the specific turbines in operation for each timestep within the planning horizon—usually spanning from an hour to a day, over a total horizon of one to several weeks. The short-term hydropower scheduling problem is challenging due to the nonlinear and nonconvex nature of hydropower production functions, and by turbine efficiencies. In regulated market settings, where energy prices can be neglected, these production functions can be modeled with high precision. However, in deregulated market settings, the inclusion of energy prices introduces additional uncertainty, further complicating a problem already affected by unpredictable natural inflows into reservoirs. This talk will focus on the short-term hydropower scheduling problem, emphasizing mathematical formulations for both regulated and deregulated market settings. Additionally, it will explore the prediction of plant operations, leveraging the extensive data available from hydropower producers.