In order to solve the problems of serious wind and solar power curtailment caused by the randomness and uncertainty of wind and solar power generation. Based on the study of pumped storage power station, high load-energy load operation characteristics and the ability to eliminate wind and solar power curtailment, this paper proposes an optimization scheduling method for multi-source complementary power plants group based on multiple temporal and spatial scales coordination. This strategy controls the multi-source complementary power generation group in three layers from the perspective of time and space, respectively. In this strategy, the multi-source complementary power plants group is controlled in three layers from the perspective of time and space, respectively. From three time scales, i.e., day-ahead, rolling and real-time, the optimal scheduling models are established with the goal of the minimum operating cost, the maximum utilization of wind and solar power storage and the minimum fluctuation of output power. Through rolling correction and real-time adjustment, the wind-solar forecasting error and the influence of load uncertainty on the scheduling plan are reduced stepwise to improve the scheduling accuracy. Combining with the specific example of Jiuquan, Gansu, it is verified that the proposed scheduling strategy can improve the economics of system operation effectively and promote the consumption of wind and solar. This work is supported by National Natural Science Foundation of China (No. 61663019).