ANN-based protection system for Controllable Series-Compensated transmission lines

This paper presents a protection system for classifying and locating faults in thyristor-controlled series compensated (TCSC) transmission lines. The proposed scheme is based on multi-layer perceptron neural networks (MLPNN). The Levenberg-Marquardt (LM) training algorithm is employed. The LM algorithm appears to be the fastest training algorithm and highly nominated for better generalized models. Three-phase power system currents and voltages at the relay location are used as inputs to MLPNN-based relay. Two neural networks are trained to address fault classification and location. Feasibility and reliability of the proposed scheme are investigated using fault data set of a typical 500 kV power system simulated in EMTPATP software package. Studied system is subjected to all possible faults at different operating conditions, including fault location, fault inception angle and fault resistance. Simulation results demonstrate the robustness and fault tolerant features of proposed protection system.