Abstract: In order to quantitatively analyze the inherent volatility of wind speed data, multifractal method is used to study the turbulence characteristics of turbulent wind field. The commonly used model of turbulent wind is introduced, based on the model wind speed data of different surface roughness are obtained. The fractal dimensions of different wind speed data are calculated with structure function dimension method, and the influence of surface roughness on the fractal dimension of wind speed data is considered. The wind speed data are analyzed with multifractal detrended fluctuation analysis method to explore the intrinsic structure of turbulent wind. And the results show that the greater the surface roughness is, the smaller the fractal dimension, the smaller the Δα (Δα is the ratio of the largest to the smallest constituent elements), and the bigger the |Δg| (Δg is the ratio of the probability of occurrence of the largest and smallest constituent elements).

Key words: turbulent wind, surface roughness, fractal dimension, multifractal