"ESTIMATION METHOD OF EVOLUTIONARY POWER SPECTRUM FOR NON-STATIONARY FLUCTUATING WIND USING WAVELET TRANSFORMS "

To further understand the time-frequency characteristics of non-stationary fluctuating wind, the relationship between time, frequency and amplitude of the evolutionary power spectrum (EPS) and wavelet transform coefficients are analyzed based on the principle of wavelet transform. The weighted wavelet function method (WWFM) is proposed to estimate the EPS of non-stationary fluctuating wind. The theoretical results are demonstrated by a series of simulated non-stationary fluctuating wind and field measurement typhoon process. The study results show that the wavelet transformation coefficients under different scales of a certain non-stationary stochastic process at a certain time are the Fourier transform of a new non-stationary stochastic process and the modulating function of the new non-stationary stochastic process is the product of the modulating function of this certain non-stationary stochastic processes and the wavelet function. The EPS of non-stationary stochastic processes can be calculated by the sum of weighted square modulus of wavelet functions at different scales and different time parameters, and the weighted coefficients are defined by the wavelet transformation results. The EPS of simulated non-stationary stochastic wind estimated by the WWFM is accordant well to the analytical results. The WWFM can estimate the EPS of non-stationary fluctuating wind effectively and the computed EPS can aid to understand the randomness of fluctuate wind further more.