Izvestiâ Vysših Učebnyh Zavedenij i Ènergetičeskih ob Edinennij SNG. Ènergetika (Feb 2023)
Symmetrical Components Digital Filters for Microprocessor-Based Protection Input Signals
Abstract
In microprocessor protections, measuring bodies are used that react to individual symmetrical components of signals or a combination of them. This makes the corresponding protection devices more sensitive. Of all the varieties of digital filters of symmetrical components, filters of the negative and zero sequences are mainly used. As for the known algorithms of their software implementation, it is advisable to use filters that form orthogonal components of the required sequence from orthogonal components of phase or phase-to-phase values of signals. Digital Fourier filters, characterized by inertia, are used as formers of the latter. For this reason, their transient modes are accompanied by the appearance of a dynamic error, which can significantly affect the functioning of filters of symmetrical components, worsening their properties. A significant reduction in this effect can be achieved by using shapers with correction of dynamic errors to isolate the orthogonal components of the input signals. They are based on non-recursive digital Fourier filters, the orthogonal components of which are subjected to software correction processing in order to obtain equivalent components characterized by fast-fading dynamic amplitude errors. Correction of the dynamic phase error is realized in the process of obtaining the resulting orthogonal components, which are a combination of Fourier components and calculated ones that have been estimated according to them. Based on the information about the amplitudes and current phases of the signals contained in the equivalent and resultant components, respectively, the resulting orthogonal components are formed, differing in minimal amplitude and phase errors in transient modes. According to these components, the orthogonal components of the signal of the negative and zero sequences in the corresponding filter are calculated. In the dynamic modeling environment of MATLAB-Simulink-SimPowerSystems, a digital model is implemented, which includes a power system, a three-phase group of current transformers, a load, a short-circuit block, as well as models of negative and zero sequence filters. The study of the functioning of these filters was carried out using two types of test actions, viz. a three-phase system of sinusoidal signals and a three-phase system of signals close to real secondary currents in short circuits. The results of the research have demonstrated that the developed digital filters of the negative and zero sequences have 1.1–1.4 times higher performance as compared with similar Fourier filters accepted as reference and that they are distinguished by improved magnitudes of dynamic indicators.
Keywords
