Abstract: The proportion of wind power in the global total power supply is also increasing year by year, and the single unit capacity of wind turbine is also increasing. At the same time, the load of the key components of wind turbine is also more and more large, and the reliability requirements of the structure and control technology are also more and more high, so the impact on the stability of the access system is more and more difficult to be ignored. Therefore, the control strategy and technology of wind turbine control system need not only the power optimization control ability, but also the load reduction control ability. The current research status of wind turbine load reduction is investigated, and the research status of model predictive control (MPC) technology is analyzed. Firstly, the wind turbine system, pitch system and drive chain system of wind turbine are modeled and analyzed, which lays the foundation for the further research of wind turbine control technology. Then, the model predictive control technology is studied deeply. Model predictive control is a kind of control algorithm to deal with nonlinear large time-delay system, which is essentially to solve an open-loop optimal control problem. In order to carry out the simulation comparative study, the variable pitch proportional integral controller of wind turbine is designed; based on the theory of model predictive control technology, the model predictive control controller of wind turbine and the multi model predictive control controller of wind turbine are designed. In the design of the model predictive control pitch controller of wind turbine, the cumulative error between the predicted output of the controlled actuator and the expected target is used as the objective optimization function; in the design of the multi model predictive control pitch controller of wind turbine, five continuous and non overlapping wind speed intervals are set, and five model predictions are designed for five different wind speed intervals At the same time, a division method of multi control objectives operation interval of wind turbine is designed to ensure that the multi model predictive control controller of wind turbine can continuously output control signals. Finally, in order to verify the load reduction ability of multi model predictive control controller for wind turbine, a real-time simulation platform based on MATLAB and TUV GL bladed is designed and written in C++. After verifying the advantages of model predictive control of wind turbine compared with traditional proportional integral controller, the nonlinear model of 2MW doubly fed wind turbine in GL bladed software is taken as the research object, and the multi model predictive control controller, model predictive control controller and traditional proportional integral controller of wind turbine are simulated and analyzed. The results show that, compared with the traditional proportional integral controller and model predictive controller, model predictive controller can effectively reduce the fluctuation amplitude of wind turbine speed, restrain the overshoot of speed, reduce the load of the transmission chain of wind turbine, restrain the fluctuation amplitude and change rate of the pitch angle of wind turbine, and reduce the operation of the pitch mechanism of wind turbine Load. It can reduce the operation cost to a certain extent and improve the overall life of wind turbine.