Recently updated on Maggio 11th, 2021 at 09:09 am

Identification of Dynamic Voltage-Current Power System Equivalents through Artificial Neural Networks

Enrico De Tuglie, Lorenzo Guida, Francesco Torelli Dario Lucarella, Massimo Pozzi, Giuliano Vimercati Politecnico di Bari CESI S.p.A. Italy Italy Abstract – This paper presents a technique for identifying a dynamic power system equivalent using Artificial Neural Network (ANN). The approach uses only measurements at points where internal (retained) and external (reduced) systems are interfaced so the equivalent is an input-output equivalent where the inputs are voltages at interconnection nodes and the outputs are currents flowing through interconnection lines. The power system is divided into two parts: internal system, also called retained system or the study system, and the external system which is equivalenced by the neural network. Given the present and past values of voltages and only past values of currents, the network can predict present current values. There is no need of the external model structure. The neural network will be interfaced with the study system through interconnection variables, i.e. voltages and currents, and will provide the dynamic behaviour of the external system. Test results will provide the robustness of the proposed methodology in developing dynamic equivalents. PUBBLICATO A5018046 (PAD – 644928)