In view of the large variation of boiler combustion conditions, the main steam pressure object model is uncertain, and it is difficult for conventional control methods to obtain the expected control effect. In order to improve the control quality, the H∞ mixing sensitivity method is used to design a robust controller for steam pressure, and SPEA2 (Strength Pareto Evolutionary Algorithm 2) is used to optimize the weighted array parameters to find a controller that meets the performance requirements of both frequency domain and time domain. The simulation results show that the designed main steam pressure control system has good robustness and dynamic quality, and avoids the traditional test stack.