Application of RBFNN to locate line to ground fault in a double end fed transmission lines using only one end data Jain Anamika1, Assistant Engineer, Thoke A.S.2, Professor and H.O.D. 1Chhattisgarh State Electricity Board, Dangania, Raipur, Chhatisgarh, India. 2Department of Electrical Engineering, N.I.T. Raipur, Chhatisgarh, India. Abstract Distance relays are used for protection of transmission lines. These relays have been reported to have problems of under-reach, over-reach & mal-operation due to the high impedance fault (HIF). Different system faults on a protected transmission line should be located correctly. This paper presents a novel application of Radial basis functions neural network to locate the line to ground fault in a double end fed transmission line using only one end data within one cycle after the inception offault. The proposed Radial Basis Neural Network Fault Locator uses fundamental components of the three phase current & voltage signals after filtering by 2nd order Butterworth filter, sampling & normalizing to learn the hidden relationship in the input patterns and is designed as a three phase relay to locate distance to fault on any one of three phases. An improved performance is obtained once the neural network is trained sufficiently and suitably, thus performing correctly when faced with different system parameters and conditions, e.g., 0∞–100Ω fault resistance (Rf), ±45 degree initial power flow angle δS, different fault incidence angles Φ etc. Top Index Terms Fault location, Fault resistance (Rf), Transmission line protection, Radial basis functions neural network. Top |