In wind power generation system, the DC-link voltage link serves as a key interface between the rotor-side converter (RSC) and the grid-side converter(GSC). Its stability has a decisive impact on the stable operation of the wind turbine system. In actual operation, fluctuations in wind speed will cause changes in the generator's output power, affecting the stability of the DC-link voltage. To address this challenge, first, this paper adopts the grid-forming virtual synchronous generator (VSG) control strategy instead of the traditional grid-following wind turbine control strategy. Then, a wind speed feedforward control strategy is proposed. Second, in turbulent wind conditions, a wind speed prediction algorithm based on the long short-term memory neural network (LSTM) is proposed to predict the changes in the effective wind speed. This allows the wind power generation system to take appropriate control measures based on precise wind speed data in advance, ultimately improving the stability of the DC-link voltage. A simulation model of a 2MW wind turbine was established using openFAST and Matlab/Simulink. The simulation results were analyzed to verify that the proposed method can improve the stability of the DC-link voltage. This research provides a new perspective for the DC-link voltage stability control of wind power generation systems.