Design and implementation of an optimal predictive controller for dual active bridge converter

Optimality controllers represent the new frontier of automatic control for several domains and applications. They are taking hold in domains such as mechanical, chemical and aerospace. This thesis presents the design, implementation, and validation of an optimality-based controller for power electronic converters, aiming to enhance efficiency, dynamic performance, and robustness. The converter on which this thesis focuses in particular, the Dual Active Bridge form the backbone of modern power systems, playing a pivotal role in applications ranging from renewable energy to electric vehicle charging. Conventional control strategies often encounter challenges in balancing multiple performance objectives, such as minimizing switching losses while maintaining output voltage regulation and ensuring system stability under varying operating conditions. To address these challenges, a model-based optimality controller is proposed that formulates control actions through the minimization of a carefully designed cost function that can take into account several variables. This controller exploits some typical power electronic methodologies such as phase shift modulation to drive the switches, but computing the optimal phase shift angles in order to track what we need to track for a certain horizon, in an MPC-like structure. By incorporating system state variables, reference tracking, and electrical and computational constraints, the proposed approach does a trade-off between efficiency and performance. Almost every metric can be tracked with the same control structure, from power transfer to current stress, load voltages and losses minimization. Extensive simulations are performed to validate the effectiveness of the proposed methodology. The results demonstrate the potential of optimality-based control in shaping next-generation power electronic systems, offering a scalable and adaptable framework that can be extended to various converter topologies and application domains