A Method for Solving Short-Term Hydro Unit Commitment Scheduling Based on Equivalent Unit Bundling during the Nonflood Season

Abstract

Short-term scheduling aims to achieve refined water allocation within a day under the framework of medium and long-term scheduling, with a focus on optimizing water resource utilization, particularly in the nonflood season. Improving the efficiency of hydropower energy conversion usually requires computation at the unit level, involving a large number of high-dimensional, discrete, and nonlinear variables, which makes the solution extremely difficult. To address such issue, this paper introduced the idea of fusion dimension reduction by bundling the same type of units within a hydropower station to establish equivalent units. Then a two-layer nested interequivalent unit and intraunit solution framework was proposed. The outer layer used an improved quantum particle swarm optimization algorithm for water allocation among equivalent units, while the inner layer employed dynamic programming to create optimal distribution tables among the units, using the bisection method for table lookup during solving. The proposed model was validated in the “water determines electricity” mode at the Three Gorges Hydropower Station, which has 34 units. Comparative analysis across different scenarios shows that (1) the proposed model improves water utilization rate and effectively increases the total daily power generation; (2) the equivalent unit strategy significantly enhances the solving efficiency of the hydro-unit commitment problem; and (3) the units under the proposed water allocation method operate well without constraint violations.