Application of a multi-reservoir dynamic collaborative optimization strategy based on extrema marginal benefits for optimizing reservoir operation

Abstract

Despite the numerous methods proposed for optimizing reservoir operation, few strategies effectively guarantee both solution optimality and efficiency. Therefore, this study proposes an extrema marginal benefits optimization method that rapidly maximizes benefits by adjusting water usage during periods of maximum and minimum marginal benefits. Given that the single-reservoir optimization model can result in a pseudo-optimal solution, this study builds on a rotational optimization model for individual reservoirs to introduce a multi-reservoir dynamic collaborative optimization strategy. This strategy targets various reservoir-boundary challenges, clearly identifying and synchronously optimizing collaborating reservoirs, thereby significantly improving the operational efficiency and optimization quality of multi-reservoir systems. The operation results of the hydropower station indicate that extrema marginal benefits optimization guarantees optimal solutions with minimal time consumption, even under high-precision conditions, where the solution time is less than one-thousandth of that required by dynamic programming. In the operation scenarios for systems with four and ten reservoirs, the multi-reservoir dynamic collaborative optimization strategy based on extrema marginal benefits reached the theoretical optimum, with the solving time never exceeding 1 s, thereby proving its efficiency and practicality.