Abstract
With the completion of the Lianghekou Reservoir, with a multiyear regulation capacity, the operation relationship of the cascade reservoirs in the Yalong River is becoming increasingly complex. In order to study an optimal operation mode of the cascade reservoirs in the Yalong River under different inflow frequencies, based on the shortcomings of the existing single reservoir operation mode and the local joint operation mode of the cascade reservoirs, this paper first proposed a global joint operation mode for the cascade reservoirs to develop the power generation potential of daily regulating reservoirs and then gave a solution method for the cascade reservoirs’ operational model based on an improved stochastic fractal search (ISFS) algorithm. Finally, taking the maximum power generation as the goal and the inflow data of five typical years as the model inputs, this paper analyzed the differences in the power generation and water abandonment results of the cascade reservoirs in the middle and lower reaches of the Yalong River under the above three operation modes. The results show that (1) compared with the stochastic fractal search (SFS) algorithm and the particle swarm optimization (PSO) algorithm, the ISFS algorithm had faster convergence speed and higher precision; (2) the global joint operation mode had a more significant optimization effect in the year with more inflow, followed by the local joint operation mode, and the single reservoir operation mode had the worst; however, the difference in the results of the three operation modes gradually decreased as the inflows gradually decreased.