Fault location using deep learning algorithm for a superconducting dc grid

With an increasing focus on using renewable energy sources and the high-power output of offshore wind turbines, it is necessary to implement systems that can handle high levels of current to simplify the process. Superconducting cables are a solution as they offer higher current densities, up to 200 times more than copper cables. A superconducting fault current limiter (SFCL) is utilized to safeguard the superconducting cables and generator. However, the utilization of superconducting cables and SFCLs can make it difficult to identify the fault location, to overcome this, a fault location algorithm that employs convolutional neural networks (CNN) is proposed in this paper. The ultimate goal of this deep learning method is to enhance the speed and accuracy of identifying faults and in turn reduce downtime and response time. The algorithm's performance is evaluated using simulations on MATLAB/SIMULINK, where various fault types and resistances will be used to validate it.