Abstract
Parameters estimation of electrical power system is relevant for monitoring, protection and control. Several methods are used to this purpose, such as Discrete Fourier Transform (DFT), Moving Average Filter (MAF), Hanning window, among others. However, some of these techniques require a number of samples within a complete number of cycles of the fundamental component in order to obtain better results. The oscillation of the system's frequency hinders the selection of a fixed window length. In this sense, in this paper is proposed a variable window length applied in a modified Hanning filter in order to optimally estimate the amplitude as well as eliminate harmonics and inter-harmonics of the power systems signals. The modified version of the Hanning filter proposed in this work enhances the magnitude response by altering the poles locations of the transfer function, achieving optimal removal of the first harmonic considering only one cycle of the signal which is unachievable by the conventional Hanning window, and the change of the window length according to the system's estimated frequency applied in this modified filter, allows for optimal removal of harmonics even when considering variable frequency.
