In order to solve the problem that the model predictive control seeks the optimal control variable, due to the unreasonable selection of the weight coefficient, the actual current cannot closely follow the given current or the predicted output voltage fluctuates too much. Attempt to apply fuzzy control to the selection of the weight coefficient. A model predictive control method based on fuzzy control dynamically selecting weight coefficients is proposed, which can send the optimal weight coefficients to the currently running model predictive control system online. And through three groups of experiments to simulation, comparison and analysis different weight coefficients, the reasonable interval of the weight coefficient is obtained. The simulation results show that the weighting coefficient based on fuzzy control has good performance. When the load change is gentle, the output voltage fluctuation will be reduced under the condition that the current can follow the given current. The improved control system improves the anti-disturbance ability, and the weight coefficient can be changed automatically according to the actual demand.