In order to reduce wind and solar power abandon caused by fluctuation of their output, this paper proposes a combined operation strategy that multi-type hydropower stations participate in wind-solar energy consumption by utilizing the regulation and storage characteristics between cascade hydropower, pumped storage and other types of hydropower. Considering the constraints of multi-type hydropower stations, reservoir capacity and transmission capacity of tie-lines, an optimal capacity configuration model of wind-solar power station coordinated by multi-type hydropower is established aiming at the minimum investment and construction cost, system operation and maintenance cost and power purchase cost and the maximum ratio of wind-solar power generation to load demand. The Extreme Learning Machine based Particle Swarm Optimization algorithm (ELM-PSO) is used to solve the model. Taking an area of Qinghai province as an example, the optimal capacity configuration scheme considering the system economy and the proportion of wind-solar power generation to load demand is obtained through simulation analysis, which verifies the feasibility and effectiveness of the proposed strategy. This work is supported by National Natural Science Foundation of China (No. 71420107027), Science and Technology Projects of Hunan Province (No. 2017CT5015 and No. 2017WK2053), Postgraduate Scientific Research Innovation Project of Hunan Province (No. CX2017B465 and No. CX2018B525), and Strategic Emerging Industries Project in Hunan Province (No. 2018GK4002).