Abstract: To explore the complementary operation of hybrid pumped storage power stations under different photovoltaic (PV) output scenarios, we developed a photovoltaic output scenario generation model based on generative adversarial networks (GANs) to characterize the uncertainty of PV output, and constructed a complementary scheduling model aimed at minimizing source load differences, maximizing system power generation, and maximizing energy storage increments of cascade hydropower stations. We optimized the model using the Fibonacci-PSO algorithm. The results showed that the hybrid pumped storage power plant, compared to the mode without such a hybrid plant, can better adjust system output to match grid load demand and utilize water power during high-water periods for power generation, increasing system power generation and resource utilization. The hybrid pumped storage plant primarily generates electricity in nights during high-water periods, while during medium and low-water periods, it generates electricity in nights and pumps water during days. This operation mode increases power generation by 4.4% to 12.4% during high-water periods and reduces the discrepancy between system output and grid load through pumping during medium and low-water periods, allowing the complementary power generation system to respond precisely to load demand. The study demonstrated that the proposed method significantly enhances the ability to incorporate clean energies.