Neural Network Two-terminal Approach for Fault Location in Transmission Line Distance Protection

This paper presents a new approa ch to locate shunt faults in power transmission lines using a two-layer Adaptive Linear Neural Network (ADALINE) in a two-end protection scheme. First, a transmission line with actual parameters is simulated in the PSCAD/EMTDC software and signals are fed to the MATLAB software to implement a real time adaptive filtering. In this approach, twoend unsynchronized measurements are used to estimate the unknown synchronization angle, fundamental frequency components using ADALINE. Finally, the distance to fault is calculated in a special process. The proposed method’s aim is to increase convergence accuracy and speed in the first period after fault inception. In this respect, the simulation results show that fault location, reactance effect and fault impedance cannot affect the proposed protection scheme.

Keywords:

distance protection , fault analysis , power transmission line , two-end measurement , tele- protection , Artificial Neural Networks