NEURAL NETWORKS FOR REAL-TIME ESTIMATION OF PARAMETERS OF SIGNALS IN POWER SYSTEMS

Fast determination of parameters of the fundamental waveform of voltages and currents is essential for the control and protection of electrical power systems. Most of the known digital algorithms are not fully parallel, so that the speed of processing is quite limited. New parallel algorithms, which can be implemented by analogue adaptive circuits employing some neural networks principles, are proposed. The problem of estimation is formulated as an optimization problem and solved by using the gradient descent method. Algorithms based on the least absolute value, the minimax, the least-squares and the robust leastsquares criteria are developed and compared. The networks process samples of observed noisy signals (voltages or currents) and give as a solution the desired parameters of signal components. Extensive computer simulations confirm the validity and performance of the proposed algorithms and neural network realizations. The proposed methods seem to be particularly useful for real-time, high-speed estimation of parameters of sinusoidal signals in electrical power systems.