A Fast-Response Method for Determining the Amplitude of a Signal in Microprocessor Automation and Control Systems with Frequency Fluctuations

Abstract

In microprocessor automation and control systems, the amplitude (effective) values of the fundamental harmonic of the input signals are widely used as information parameters of the controlled quantities. They are most often determined by samples of one or a pair of orthogonal components of the signal, for the formation of which digital Fourier filters and their modifications are mainly used. At the rated frequency in the power system, these filters ensure reliable reception of the signal amplitude without additional error. If the frequency deviates from the rated one, the number of samples per signal period is not an integer and the discretization becomes asynchronous. As a result, a corresponding error appears in the amplitude of the signal, and its change becomes oscillating. With minor frequency fluctuations in the normal mode, the amplitude error is insignificant. However, in abnormal situations, the frequency can have significant variations. At the same time, in critical situations, failure of automation and control systems, as well as incorrect operation of their functional algorithms, cannot be excluded. Known methods for determining the amplitude of a signal with frequency fluctuations provide a solution to the existing problem, but they are characterized by a slow response. The proposed high-response method for determining the amplitude during frequency fluctuations is focused on using as initial information samples of instantaneous values of the cosine orthogonal component of the signal, which are formed using an appropriate digital Fourier filter. Based on these samples, the dynamic cosine and sine of the angle of one sample are calculated, the use of which in calculating the amplitude ensures its independence from frequency. Processing of the received amplitude with an amplifying element with a nonlinear coefficient makes it possible to achieve acceptable performance. The effectiveness of the proposed solution was evaluated by a computational experiment using a digital model implemented in the MATLAB-Simulink dynamic modeling environment. In this case, both sinusoidal input signals and complex ones, close to the real secondary signals of measuring transformers, were used as test actions. As a result of the research, it was found that the proposed method for determining the amplitude during frequency fluctuations has a performance at the level of a quarter of the period and provides effective elimination of frequency error both in load modes and in damage modes.