Abstract of Thesis presented at COPPE/UFRJ as a partial fulfillment of the requirements for the degree of Master of Science (M.Sc.)

Classfication and Fault Distance Location for Series Compensated transmission Systems Using Artificial Neural Networks

Diego de Souza de Oliveira

January/2016

Advisors:

Sebastião Ércules Melo de Oliveira Luiz Pereira Calôba

Department:

Eletrical Engineering

      This work has as main objective the performance evaluation of a detection, classification and fault location scheme in transmission systems with series compensation using Artificial Neural Networks (ANN), starting from the analysis of the post-fault three-phase signals of voltages and currents obtained from both line terminals. The RNA architecture was set from training, validation and testing data sets contained in database generated through simulation of short-circuit situations on a high voltage transmission system. The simulations were performed using MATLAB software and modeling of elements carried out in its toolbox "Simulink" from models of system elements available in the power system library "SimPowerSystems".
      Results are presented highlighting the difficulties of conventional distance protection schemes from the point of view of the series compensated systems. Emphasis is given to the effects associated to the subsynchronous resonance and DC current components, to the action of series capacitors overvoltage protection using metal oxide varistor protection (MOV) on the impedance measurement in conventional relays and the impacts of these effects on the RNA performance.