Selecting wavelets for damping estimation of ambient-excited electromechanical oscillations

This paper discusses aspects related to a wavelet transform and random decrement technique based method developed for estimating power system electromechanical oscillation damping from ambient-excited oscillations. The main focus of this paper is in finding the optimal wavelet functions for the damping estimation method. General selection criteria for the wavelet functions are defined, the selection criteria are specified quantitatively for the Nordic power system and the wavelet functions fulfilling the criteria are presented. Damping of a simulated Nordic power system case is analyzed with the damping estimation method and with different wavelet functions. The damping estimation results show that when the mother wavelet is selected according to the specified criteria, the damping estimates are better than by using other wavelets.