Известия высших учебных заведений и энергетических объединенний СНГ: Энергетика (Feb 2025)
Correction of Amplitude and Phase Errors of the Signal in Microprocessor Automation and Relay Protection Systems when Frequency Changes
Abstract
In microprocessor automation and relay protection systems, amplitude (effective) values and phase shifts of input signals are widely used as controlled parameters of electrical quantities. In most cases, they are determined by samples of one or two orthogonal components of the main harmony of these signals. In existing microprocessor systems, non-recursive digital Fourier filters are mainly used to form them. At a nominal frequency in the power system, the orthogonal components highlighted by these filters do not introduce additional errors into the amplitudes and phase shifts determined by them. In modes with frequency deviation from the nominal value, the number of input signal samples per period becomes a fractional number, and sampling turns into asynchronous one. For this reason, corresponding errors appear in the amplitude and phase of the signal. The main issue of their correction is the direct or indirect estimation of frequency. In this article, an indirect estimation of the instantaneous frequency is realized by the dynamic cosine of the angle of one sample, which is calculated from three adjacent numerical values of the cosine orthogonal component of the signal. The use of instantaneous frequency information in determining amplitudes and phase shifts allows for full correction of the corresponding errors. It should be noted, however, that in transient modes, due to the influence of various factors, the dynamic cosine is determined with large errors. This makes it impractical to correct amplitude and phase errors in these modes, limiting its use only to steady-state modes. For these modes, a functional algorithm for correcting the amplitude and phase errors of signals in microprocessor automation and relay protection systems when the frequency deviates from the nominal one is proposed and investigated. The results of the performed investigations showed that the developed correction algorithm provides almost complete elimination of the manifestation of amplitude and phase errors in steady-state modes in the frequency range of 47–52 Hz.
Keywords