In order to solve the problem of a balanced water resources operation more effectively, the improved particle swarm optimization (PSO) algorithm is applied to the optimization of water resources and a mathematical model is constructed of water resources to optimize the operation of cascading hydropower plants. The objective function of the model is to solve the optimal output process of each hydropower plant so as to maximize the total power generation during the operation period. The constraints of the model are water level, output, discharge, the water balance of the reservoir, and the water flow relationship between the reservoirs. The particle swarm optimization algorithm is used to optimize the water resources of cascade hydropower plants. The objective function is taken as the fitness function of the particle swarm and the particle flight speed is dynamically adjusted to the ultimate optimal point. The time dimension of optimal scheduling is taken as the particle space dimension to realize the calculation of the multi-stage process of the scaling period. On this basis, from the two aspects of local convergence improvement and PSO, we improve the resource balance scheduling process of the particle algorithm and improve the balanced scheduling of water resources. The results show that the power generation of cascading hydropower plants is 4.52 % higher than that of the conventional operation when using the PSO algorithm; the average annual power generation of cascading hydropower plants is increased by 2.0 %, the guaranteed output is increased by 1.9 %, and the average annual waste water is reduced by 1.6 %. The guaranteed degree of water supply of the reservoir during the operation period is reduced and the storage capacity of the reservoir at the end of the year is increased by 50 %. The feasibility and effectiveness of this method in the optimization of water resources balance are verified, which shows that this method can provide a new idea for the balanced operation of water resources.