Award

Hepburn and Sons LLC, teamed with Florida State University Center for Advanced Power Systems (FSU CAPS), is proposing to develop a machine learning (ML)-based fault protection relay system, complete with necessary sensors, for 1kV, 6 kV, and 12 kV zonal Medium Voltage Direct Current (MVDC) shipboard power systems (SPS) with the objective to implement overcurrent and differential/directional protection in both breaker and breakerless architectures. MVDC SPS architectures present unique challenges and opportunities for protection due to the low system inertia, high probability of backfeeding through low impedance buswork, and imprecise zonal definitions due to the lack of transformers.  It is critical for the maintaining of SPS integrity to quickly isolate a faulted piece of Navy equipment, and the use of advanced methods that identify the electrical zone and shipboard compartment containing the fault will facilitate rapid repair responses and enable safe and continuous operation of the system for the warfighter.  Our proposed relay system will leverage emerging machine learning technology to quickly characterize fault signatures in zonal architectures of interest.  This solution will provide the benefit of simplified, high speed, fault identification and localization to then coordinate and trigger appropriate protection responses in a proper sequence upon fault initiation.