Abstract: As extreme weather events become more frequent due to climate change, reservoirs face growing challenges in flood control scheduling. Taking the Xiaogang watershed in Songyang County, Zhejiang Province as the research area, hydrological forecasting models, reservoir scheduling models, hydrodynamic models, and flood risk analysis models were constructed. Based on actual reservoir scheduling and historical flood cases, the forecasting models were calibrated and validated. The research results indicate that the model can effectively improve the accuracy of flood forecasting and optimize reservoir scheduling strategies to reduce the impact of flood disasters on downstream areas. This study not only provides scientific technical support for reservoir flood control and scheduling, but also provides new theoretical support and practical guidance for improving the emergency response and disaster prevention capabilities of water resource management.