Wind speed measurement method using ultrasonic sensors with stationary wavelet transform

A method for measuring speed wind, based on echo detection using ultrasonic transducers with stationary wavelet transform is presented. The stationary wavelet transform (SWT) can be obtained by modifying the basic scheme of discrete wavelet transform (DWT). Two filters are applied to produce the output two sequences at each level. Since decimation is not obtained at the exit of each level, sequences of length equal to the original signal. But the filters are changed at each level adding zeros. After obtaining the SWT coefficients, noise removal will be performed using the soft-threshold rule due the coefficients still include the effects of noise. The measured ultrasound signals are processed using the stationary wavelet transform. The accuracy of the extracted information from the processed signal depends on the adopted mother wavelet. To select the best mother wavelet for signal processing, a criterion based on Shannon's entropy is used. The influence of uncertainties associated with additive noise and attenuation of the ultrasonic signal on the wind speed measurement uncertainty is analyzed.