Abstract: The regulation of water and sediment in the Upper Yellow River is studied in this paper. To relieve the secondary suspended river that has formed in Inner Mongolia, desert valley reaches in the Upper Yellow River were studied and Longyangxia and Liujiaxia cascade reservoirs were chosen as subjects. Single objective models of sediment discharge maximization and generated energy maximization and a multi objective model were established, and solved by self-iteration simulation algorithm, successive approximation dynamic programming(DPSA), and an improved non-dominated sorting genetic algorithm(NSGA), respectively. Then initial schemes, routine schemes, optimization schemes, and joint optimization operation schemes were set. The results demonstrated that the total scouring sediment of joint optimization schemes was 38 million tons, and power generation of the cascade was 14.8 billion kW·h, with a small loss of power generation in exchange for a substantial increases in sediment. This indicated the great effort required for sediment regulation and the joint operation schemes was recommended as the best option. The results quantified the water and sediment regulation effects and transformation law between objectives are shown. Moreover, the achievements provide a basis for decisions regarding the regulation of water and sediment in the Upper Yellow River. They may have important applications and practical significance in this type of regulation.