Bi-Level Energy Management and Design Optimization for Hybrid Marine Power System Applications

This paper presents a two-level nested optimization methodology for energy management and component sizing optimization in hybrid marine energy systems. The approach combines mixed-integer linear programming (MILP) for highresolution energy management system (EMS) optimization and Bayesian optimization for overall system design. By integrating high-resolution operational data and low-resolution design data, the proposed framework enables optimal integration of battery and supercapacitor systems in maritime vessels. The methodology is applied to real-world case studies, demonstrating its effectiveness in supporting the adoption of innovative battery (such as semi-solid state Li-ion cells) and supercapacitor chemistries. The results show significant potential to reduce fuel consumption and emissions, contributing to greener and more efficient shipping. This work highlights the applicability of bi-level optimization as a decision-making tool to advance sustainable marine energy systems.