Numerical Simulation Of Stochastic Fluctuating Wind Field

Air movement in surface layer is a kind of random turbulent motion. Usually, the wind speed which a point in space is considered the sum of average wind and fluctuating wind when we study the characteristics of turbulent motion. The average wind speed does not change with time and the fluctuating wind that has reflected turbulence characteristic of the atmospheric boundary layer. It is intensity and direction change with the time. Their period are only in seconds to tens of seconds.So the basis of study the atmospheric turbulence movement is accurate and efficient simulation of fluctuating wind. This paper simulate the fluctuating wind field in the surface layer and the content consists of several parts.Firstly, this dissertation summarized the current situation of wind field simulation and the essential characteristic of wind field. The fluctuating wind is usually modeled as a stationary Gaussian stochastic process.The procedure which was programmed separately based on linear filtering method and harmony superposition method and inversed dispersed fourier transform method to simulate the along and vertical fluctuating wind which a point in space. Then compared the target power spectrum and the simulation power spectrum and calculated the auto-correlation function with the time interval of 0 second. The simulation results show that these compare are in good agreement and the three simulation method are feasible and effective.Secondly, analysis the advantage and disadvantage of three kinds of simulation methods.Then use the linear filtering method to simulate relevant fluctuating wind. The calculation program was compiled according to the basic algorithm. Then simulate stochastic fluctuating wind field of surface layer under 10 meters based on target power spectrum. We can observe that the simulation of along and vertical fluctuating wind speed time-history curve randomness is well, the target power spectrum and the simulation power spectrum conformity is good. Then through calculate the cross-correlation function we get the following conclusion that the adjacent points coherence is high, when their distance is farther, the coherence is weaker. These conclusions are in conformity with the existing correlation test.Finally, calculate the along fluctuating wind turbulence intensity and the turbulence integral scale and fitting out the expression which change with the height distance and surface roughnessbased on the simulation results.