In allusion to the problem of integrated scheduling optimization in a variety of new energy power generation, the paper integrates the wind power, photovoltaic power, concentrating solar power and thermal power generation as the Multi-Source Power System (MSPS), and a two-stage stochastic scheduling optimization method based on robust stochastic optimization theory is proposed. First, this paper introduces the basic structure of MSPS and establishes the output power models of MSPS. Then, the scenario generation and reduction frame based on the interval method and the probability distance is proposed for dealing with the randomness of new energy output power. On this basis, a two-stage stochastic scheduling optimization model for MSPS is established. Simultaneously, this paper transforms the constraints containing random variables into the constraints reflecting the decision maker’s attitude of bearing systematic risks based on robust stochastic optimization theory, and then establishes MSPS stochastic scheduling optimization model including two robust coefficients. The simulation proves that the model can reduce the risk of power shortage and provide scheduling optimization decision-making foundation for different-risk-attitude system manager. This work is supported by Science and Technology Project of State Grid Corporation of China (No. SGGSKY00FJJS 1800140) and National Natural Science Foundation of China (No. 61663019).