Improving Flood Control Optimal Operation of River-Type Cascade Reservoirs through Coupling with 1D Hydrodynamic Model

River-type cascade reservoirs are hydraulically connected and have a dynamic reservoir capacity. Previous studies on the flood control optimal operation (FCOO) of cascade reservoirs have considered the hydraulic connections among them, but the scheduling calculations were typically based on a static capacity approach, which is inadequate for river-type reservoirs. To address this issue, an improved model for FCOO in river-type cascade reservoirs is proposed by coupling a one-dimensional hydrodynamic model. The improved model depicts both the hydraulic connections among reservoirs and the transformation of reservoir states using one-dimensional unsteady flow simulations. Additionally, a term accounting for the temporal variation in wedge-shaped reservoir capacity is incorporated into the traditional objective function. Parallel dynamic programming successive approximation is combined with improved discrete differential dynamic programming to solve this model. The applicability of the improved model is demonstrated through its application to the Xiangjiaba Reservoir (XJB) and Three Gorges Reservoir (TGR). Compared with the static capacity method, the improved model reduces the maximum flow at Lizhuang station by 266 m3/s for the XJB, but increases the dam-front water level by 0.12 m and the maximum flow at Zhicheng station by 3,159 m3/s for the TGR. This is because the improved model considers the floodwater retention at the tail of the reservoir and from previous flood processes. The extra term in the objective function is found to contribute to the proposed model's superior performance. This research provides reliable information for refined FCOO simulations in river-type cascade reservoirs.