Non-stationary Fluctuating Wind Velocity Spectra Based On Wavelet Transform

Wind speed is assumed to be stationary stochastic processes in the existing studiesof structural wind-resistant design. The factors affecting wind speed is complex and partof wind speed data is the non-stationary stochastic process in the process of wind speedmeasurements. Wind non-stationary is more obvious especially in complex terrainplaces or under adverse weather conditions, such as wind environment, typhoons,hurricanes and so on. There is no doubt that some errors will be caused without theconsideration of wind non-stationary in the process of wind design.The non-stationary characteristic of turbulent wind was analyzed based on asample of measured wind speed data of one mountain in Chongqing. The study methodof wind speed non-stationary model was proposed through analyzing the non-stationaryfluctuating wind under complex terrain, and then the method to estimate the wind powerspectrum of non-stationary model was summarized through wavelet transformation.Based on the evolutionary spectrum theory, stationary fluctuating wind speedtime-history was simulated generated through the method of harmonic superposition,and then the numerical simulation of non-stationary fluctuation wind was made by usingthe modulation function. The wind speed spectrum estimated by the short-time Fouriertransform was used in this paper, and it was compared with the theoretical value.The data based on the db wavelet, sym wavelet, coif wavelet which used to extractthe trend and compare precision of the trends was compared with the theoretical valuein this paper in order to know the accuracy. Then, controlling the quality of themeasured data, excluding dead pixels and testing the non-stationary of the upper dataafter removing the trend of non-stationary samples, the non-stationary characteristic ofmeasured complex mountainous terrain was figured out.The transforming relationship between the traditional concept of power spectrumand the wavelet coefficients was calculated in this paper. The accuracy of powerspectrum of non-stationary fluctuating wind estimated by wavelet transform wasverified by comparing with the theoretical value. Then, the wavelet transform was usedin this paper to estimate the fluctuating wind of the measured non-stationary wind speeddata, additionally, the analyzing method of non-stationary fluctuating wind based onwavelet transform was proposed in this paper.