Efficient simulation of wind fields based on the factorization of wavenumber‐frequency joint spectrum

Abstract

AbstractThe fast Fourier transform (FFT) technique is usually adopted in the wind field simulation based on the wavenumber‐frequency joint spectrum. However, inefficient multi‐dimensional FFT in both frequency and wavenumber domains are directly implemented in the existing literature. In this paper, a novel factorization method of wavenumber‐frequency joint spectrum is proposed to enhance efficiency. For the one‐spatial dimension wind field, the wavenumber‐frequency joint spectrum is decoupled by employing the proper orthogonal decomposition technology, and it is proposed to factorize the double‐index random phase angle into two independent random phase angles. Then, the previous 2D FFT can be replaced by 1D FFT. Furthermore, the proposed factorization method is extended to the two‐spatial dimension wind field, considering two wavenumbers and triple‐index random phase angle are introduced, the wavenumber‐frequency spectrum and triple‐index random phase angle are decomposed twice in succession. Thus, the undesired 3D FFT can be avoided. Simulations of a one‐spatial dimension and a two‐spatial dimension wind fields are used to demonstrate the accuracy and efficiency of the proposed method by comparing it with the existing method involved in multi‐dimensional FFT.