A Residual Attention Aware Network Based Faults Characterization of a Power System with SFCL

Faults in power systems are responsible to increase the current abruptly that gradually damages the protective equipment and loads. Although the use of a superconducting fault current limiter (SFCL) decreases the level of fault current, it considers 2 cycles for detecting the fault where the conventional circuit breaker takes it up to 5 cycles. In this paper, we propose a deep learning network to detect and classify the fault with a fraction of the cycle for the rapid restoration of faulty phases. The decayed current signals using the SFCL are considered to measure the effectiveness of the proposed learning network. The obtained result shows that the proposed learning approach can detect and classify the fault within the fraction of one cycle.