The Yellow River Protection Law explicitly calls for improving the water-sediment regulation system and enhancing sediment transport capacity. The late stage of pre-flood water-sediment regulation is a critical period for large-scale sediment flushing and capacity recovery at the Xiaolangdi Reservoir. Optimizing the scheduling process of the Sanmenxia Reservoir can effectively provide subsequent driving force for the Xiaolangdi Reservoir. Extensive research has been conducted by scholars both domestically and internationally: research on reservoir scheduling has evolved from single-objective scheduling to multi-objective collaborative optimization, while sediment transport calculation research has diversified into various data-driven models. Despite these advances, research specifically focused on the pre-flood water-sediment regulation period remains insufficient. In particular, the scheduling process at the Sanmenxia Reservoir lacks accurate methods for calculating sediment discharge, and there is limited research on simultaneously enhancing subsequent driving force while balancing hydropower generation and sediment transport benefits.This study focuses on the Sanmenxia Reservoir. Based on water and sediment data from the Tongguan and Sanmenxia hydrological stations from 2003 to 2023, an XGBoost algorithm is used to construct a sediment discharge prediction model. Additionally, a multi-objective optimization scheduling model for the Sanmenxia Reservoir is established using the NSGA-II algorithm to solve scheduling schemes under different inflow water and sediment conditions. The corresponding optimal solutions under these conditions are identified using the VIKOR multi-attribute decision-making method. The results show that the sediment discharge prediction model developed in this study achieves high accuracy and can be effectively used for calculating sediment discharge. Furthermore, the multi-objective optimization scheduling model outperforms the actual scheduling process under typical conditions during late June to early July of the pre-flood period. It not only increases the number of days with discharge exceeding 2600 m3/s but also significantly enhances the daily average discharge during the sluicing period, leading to an overall improvement in outflow during the sluicing period. This effectively alleviates the problem of insufficient subsequent driving force at the Xiaolangdi Reservoir caused by limited natural inflows during the pre-flood stage.