ML 기법을 적용한 HEPS용 전력 제어시스템 알고리즘 연구

Abstract

Pollutants emitted through ships have a major impact on global warming and climate change. And the IMO(International Maritime Organization) amended the Convention on the Prevention of Pollution from Ships to limit greenhouse gas emissions through ship energy efficiency regulations. To this end, the HEPS(Hybrid Electric Propulsion System), which meets the requirements and minimizes fuel consumption, is providing an optimal solution. However, the propulsion load and pulse load of the ship will have a negative effect on the onboard power system. And a stable power management strategy that can manage power distribution between different power sources in real time as well as reduce fuel consumption is needed. Therefore, in this paper, we propose a LCS(Load Control System) algorithm for systematic power supply of generators and batteries. To this end, the generator and battery capacity were calculated, and the ratio of operation in the optimal efficiency range was examined when the LCS algorithm was applied by creating a load scenario according to the operation mode. And, if the abnormal state in the system interlocked with the LCS is identified, the loss of operation can be minimized. To this end, based on the flight data obtained through the LCS, basic research necessary to implement an ML(Machine Learning) based anomaly detection function is performed.