Abstract: Due to the influence of microtopography and micrometeorology,the randomness characteristics of wind speed in wind farms are complex. To accurately describe the randomness characteristics of wind speed,a feature modeling method based on component dependency is proposed. Firstly,the randomness of the wind speed series is extracted,and the wind speed is decomposed into several modal components with different frequencies by using the variational mode decomposition（VMD）. With the serial autocorrelation coefficient（AC）as the index,the wind speed component is divided,and the fluctuation component and randomness component of wind speed are obtained. Then,considering the dependence of the wind speed randomness component on the fluctuation component,the normal distribution is used to describe the randomness feature under different wind speeds,and a wind speed randomness model based on component dependency is established. The effectiveness of this method is verified by the operation data of a wind farm in Zhangjiakou,North China. The experimental results show that this method can better reproduce the random characteristics of wind speed series.