Abstract: Combining flow and sediment dynamics with optimization theory, we developed a technology of sediment control for cascade reservoirs integrating simulation module, dispatching module and decision module.Firstly, we improved the coefficient of saturation recovery and the sediment carrying capacity of non-uniform sediment, and on this basis built the simulation module of water/sediment transport in rivers, lakes, and reservoirs by integrating the water/sediment models for cascade reservoirs in upstream Yangtze River, for complex river-lake network in mid-and downstream Yangtze River and the 2D model for typical reaches. Subsequently, we constructed the objective function of sediment’s optimal regulation in consideration of flood control, power generation, navigation, and long-term use, and further established the optimal sediment-dispatching model. Furthermore we employed the BP neural network in association with the pre-constructed sediment information base to fit the prediction for the sedimentation and scouring, and present an improved grey-target evaluation method based on non-inferior solution set. Taking the Three Gorges Reservoir as an example, we finally put forward the dynamic operation scheme of “storing clear water and discharging muddy water” in flood season and the long-term staged sediment control strategy.

Key words: sediment regulation, cascade reservoirs, sediment movement, long-term use, multi-objective water-sediment mathematical model, storing clear water and discharging muddy water, Yangtze River basin