Long-Term Feedback Enabled Energy Management Control Framework for Ships with Hybrid Power Supply

Long-Term Feedback Enabled Energy Management Control Framework for Ships with Hybrid Power Supply

ANTONOPOULOS, SPYROS (TU Delft Mechanical, Maritime and Materials Engineering)

Visser, K. (mentor)
Ramirez, Andrea (graduation committee)
Reppa, V. (graduation committee)
van Biert, L. (graduation committee)
Kalikatzarakis, Miltos (graduation committee)

Delft University of Technology

Marine Technology

2020-07-13

In the present study, a Model Predictive Control (MPC) - based, Energy Management Strategy (EMS) is proposed, aiming at the minimisation of the fuel consumption of vessels with hybrid propulsion and power supply. This approach is capable of utilising online information regarding realistic future predictions of the propulsive power demand, addressing the current gap in the literature regarding the implementation of MPC EMSs onboard ships, and pragmatic future mission information utilisation. To the author's best knowledge, MPC for fuel consumption minimisation has not yet been implemented at this timescale in maritime propulsion plants with hybrid power supply. Furthermore, online mission information from the Integrated Bridge Systems (IBSs) and the governor's human input cannot be utilised by EMSs onboard ships, due to the lack of a mathematical pathway to enable such an information feedback flow. Increasingly powerful processing units in ship controllers and a wider application of batteries in marine powertrains motivated such a computationally heavy and, simultaneously, highly fuel-efficient control approach. The proposed controller framework has been tested and verified in a case study using a Mean Value First Principle (MVFP) propulsion plant model of a hybrid naval vessel with batteries in Simulink®, provided by DAMEN® SNS. By using provided mission data, realistic mission profiles were generated, in which the proposed controller is validated against perfectly and piecewise, non-perfectly informed Dynamic Programming (DP) solutions in an information barrier scheme, yielding close to optimal performance, and a fuel consumption reduction of up to 3.5% when compared to any non-long-term feedback enabled controller.

Energy Management Strategy
Model Predictive Control
Hybrid Power Supply
hybrid electric propulsion
Naval Ships
Integrated Bridge Systems
Long-term Information Utilization
Control Framework
Feedback Enabled

http://resolver.tudelft.nl/uuid:3d7c42f0-8178-4ab6-8a76-95efea42f7bf

2022-07-13

Student theses

master thesis

© 2020 SPYROS ANTONOPOULOS