Lead PI: Dr Ahmed Elkafas, University of Strathclyde

Summary:

TWIN-POWER aims to develop an advanced digital twin and optimization platform to support design and operation of sustainable power and propulsion systems for UK and global fleets. The project integrates deep learning–based digital twin, trained on high-frequency operational data, to generate time-dependent power demand and fuel consumption profiles under realistic operating conditions.

These outputs feed MATLAB-based multi-criteria feasibility platform evaluating innovative power technologies against technical, economic, environmental, and regulatory criteria. A time-dependent MILP optimization framework then identifies optimal hybrid power system configurations and operational strategies. TWIN-POWER enables data-driven, stakeholder-informed decisions to improve performance, fuel optimal use, and decarbonisation outcomes.