Application of wavelet denoising of improved threshold function to vibration signal analysis of hydroelectric units
Jin Wang · Journal of Hydroelectric Engineering · 2012
Threshold denoising in wavelet domain is a traditional algorithm for noise reduction of hydro turbine vibration signals.This paper puts forward a denoise analysis method of improved threshold function for turbine vibration monitoring to solve the traditional method's conflict between noise pollution control and signal details retaining.Signal sampling and denoise calculation are implemented with a code of Labview and MATLAB languages.A comparison of the denoise capability of simulated signals verifies that this new method gives better RMSE performance and RSN gains than the traditional method.A detailed comparison of retained vibration characteristics of the original signals with the denoised signals by different denoise methods shows that the new method is effective and superior in retaining each characteristic component to the traditional function.