<![CDATA[The increasing number of fossil fuel-powered vessels in Indonesia poses serious challenges related to energy efficiency and exhaust emissions. This study aims to design and implement a hybrid propulsion ship speed control system based on the Neural Network (NN) method to improve fuel efficiency, speed stability, and system safety. The system was developed on a trimaran ship prototype that combines a Brussles Directional Current electric motor and an Internal Combustion Engine (ICE). The Neural Network method is designed with a 5-8-3 architecture as an adaptive controller, processing actual speed and RPM inputs to generate real-time motor control signals. Tests were conducted statically and dynamically at five speed set points under current conditions of ±0.57 m/s and waves of 2–15 cm. The test results showed that the application of Neural Network control improved speed stability, with a deviation of <0.5 km/h at the low set point, and significantly reduced RPM fluctuations. Fuel efficiency also improved, especially at high speeds; the increase in consumption was only ±3.42 ml with control compared to ±43.6 ml without control. In addition, the system is equipped with overcurrent protection that effectively prevents component damage, as proven after initial trial incidents. Overall, the results of this study demonstrate that the integration of Neural Network methods into hybrid propulsion systems has great potential for the development of energy-efficient, stable, and environmentally friendly smart ships.]]>
Copyrights © 2025
