Conclusions
- The development of a hybrid propulsion system model using fuel cells and batteries is explained. It is emphasized that the system is optimized under a multi-objective setting.
- How the energy management problem is formulated using a Markov decision process is explained. It is highlighted that this formulation represents a generic energy management framework applicable to hybrid propulsion systems.
- How deep networks are used to solve the optimal energy management problem is explained. The application of reinforcement learning to control system operation is focused on.
- How the reinforcement learning energy management framework can be generalized beyond marine applications is explained. It is noted that the same framework can be extended to other sectors.
Updated on Jan 18, 2026