Shunt Reactor Switch-On Phase Estimation by Adaptive Structural Analysis Methods: Fid-Bau Portal

Shunt Reactor Switch-On Phase Estimation by Adaptive Structural Analysis Methods

Aleksandrova, M. I. / Antonov, V. I.

Anew approach is proposed to solve the problem of determining the switching moment in a shunt reactor controlled switching device, based on estimating the ratios of the initial value of the free component and the amplitude of the steady-state component of the controlled switching by the methods of adaptive structural analysis. The tool for recognizing the signal structure in adaptive structural analysis is a digital structural model. The roots of its characteristic equation determine the characteristic parameters — the frequencies and damping coefficients of the signal components. The complex amplitudes of the signal components are determined in the component model of the signal, formed on the basis of the effective core components of the structural model. Two methods are described. The first method uses the classical adaptive structural model. In this case, the signal of the shunt reactor switching current is processed by a single adaptive filter, and as a result of component analysis, estimates of the free and steady-state current components are obtained. The second method is based on the application of a hybrid structural model. Its non-adaptive part consists of a fundamental harmonic suppression filter of the current. The adaptive filter in the method is adjusted only for the aperiodic component, which in turn makes it possible to reduce the order of the model. The amplitude of the fundamental harmonic is estimated by the orthogonal components of the signal current. It is shown that the estimates of the shunt reactor switching phase obtained by both methods are comparable and have high accuracy. The use of these methods to determine the switching moment provides the possibility of implementing the controlled switching function in automation devices without a large sampling frequency of the analog-to-digital converter (ADC). In this, preference must be given to the estimation method of the hybrid structural model, since its implementation requires less computational resources.