Project Summary

The project considers the modelling and control design of a power management system to optimize the energy used in a zero-emissions marine vessel. The distribution of energy between power sources, storage systems and actuators will be controlled using an advanced model based optimal control design approach.

Project Achievements

Activities Task 1: • Analysis of fuel cell vessel specifications • Design and validation of the vessel simulator Task 2: • Design and validation of the MPC-based EMS • Design and validation of the AI predictor • Integration and validation of MPC-based EMS with AI predictor

Conclusions

Relative to a to a baseline rule-based (RB) control a basic MPC provides real improvements, since degradation can be tuned via the cost weightings. The MPC with AI, results in even better fuel cell degradation and hydrogen consumption. With respect to a baseline rule-based (RB) policy, the MPC with AI prediction reduces fuel cell degradation by 99% and hydrogen consumption by 16%. With respect to a basic MPC the MPC with AI prediction reduces fuel cell degradation by 62.5% and reduces battery stress by 2.32% and hydrogen consumption by 2.5%. The battery degradation is somewhat worse for both MPC designs but economically this is not so important as the improvement in fuel cell degradation.

Next Steps

Initial contacts have been made with potential future partners and customers. The next steps are: • A prototype should now be produced which requires investment we must secure, and testing in controller in the loop simulations ideally followed by ship trials. • Futher contacts with potential user companies and presentations of the results. • Submission of a paper to attract support at an exhibition, workshop or conference. • Include the material in ISC training courses that can be provided at company premises.