An innovative finite state automata based approach for fault direction estimation in transmission lines

In this work finite state automata or finite state machine based directional protection scheme is proposed for transmission lines. Phase angle of positive sequence current is used as input to finite state automata based fault direction estimation module. Finite state automata is used as fault pattern recognizer which estimates the direction of fault. The output of the proposed FSM based scheme will be '−1' for reverse section faults and '1' for fault in primary section faults. The performance of the proposed technique is evaluated using data simulated for variation of fault type, fault inception angle, fault location, power flow angle, reverse power flow and fault resistance. Accuracy of the method is found to be 100% from all 11,500 fault cases. Proposed technique does not use voltage unlike conventional directional relaying schemes due to which there is no issue regarding close-in fault detection. The reach setting of the proposed method is up to 99.9% of line length which has advantage over conventional relaying schemes which have reach up to 80-85% of line. Although proposed method is a pattern recognition based technique, it does require an extra training module unlike artificial neural network to estimate the direction correctly. The proposed technique is effective because it do not require any training and the computation complexity is very less as compared to training based algorithms. Proposed method is also tested in an existing power system network of India, which shows accurate result in estimating fault direction.