Pulsed-Power Load Monitoring for an All-Electric Ship: Utilizing the Fourier Transform Data-Driven Deep Learning Approach

Abstract

Future Naval Advanced Weapon and sensor loads require high power density power electronic distribution systems (PEDs) to be viable. The mission-defined electronic loads, such as advanced radar and pulsed-power weapons, tend to have an irregular power draw, consume a significant amount of power, and would require significant additional generation equipment well beyond the existing power system. A legacy low-voltage ac power grid on a destroyer is already at capacity and because space is at a premium, it is unlikely that any new power-hungry pulse load could be included without a significant paradigm shift. Furthermore, most of the potential power within the vessel is locked to the mechanical drive system and not readily usable as electrical power. As such, the all-electric ship is the intended option to unlock the mechanical power for various high power pulse loads.