Dimensionality Reduction Method of Dynamic Programming under Hourly Scale and Its Application in Optimal Scheduling of Reservoir Flood Control

In flood control scheduling of reservoirs, the usual conventional scheduling fails to obtain the optimal solution to the problem. The dynamic programming method is applied to the field. However, the problem of ‘dimensional disaster’ restricts its application. To solve these problems, the improved DP is constructed according to the principle of period inflow and water balance, which includes a variable dispersion mechanism of the retraction space and a calculation method for the smallest discrete points of each discrete range. Taking Dongjiang Reservoir as the research object, based on the maximum peak clipping criterion and the maximum flood control safety guarantee criterion, the improved DP is used for optimization. The results find that when the former criterion is used for scheduling, the discharge flow processes of the two floods are more uniform than conventional scheduling. When the latter criterion is used for scheduling, the minimum water levels of the two floods are lower than the conventional scheduling. As such, it is found that in the two flood dispatches, the calculation time after the first flood DP dimensionality reduction processing is reduced by about 65%, and the second flood is reduced by about 59%, which greatly improves the calculation efficiency of the DP.