Advanced Modelling, Optimized Design and Innovative Control Strategies for Dual Active Bridge DC-DC Converters for Efficient Electric Vehicle Fast Charging

The rapid growth of Electric Vehicles (EVs) as a key solution to reducing greenhouse gas emissions requires the development of efficient, robust, and versatile fast-charging infrastructure. This research focuses on the design, control, and optimization of Dual Active Bridge (DAB) DC-DC converters, essential for high-power EV charging stations due to their soft-switching capabilities and adaptability to various battery chemistries. A key challenge is ensuring smooth transitions between charging stages, such as Constant Current (CC) and Constant Voltage (CV) modes, vital for optimal battery health and charging efficiency. The study begins by investigating the dynamic operation of DAB converters, introducing an improved Fourier series-based Generalized Average Model (GAM) that accounts for parasitic resistances in output filter capacitors. This enhances simulation accuracy, providing a more realistic depiction of the converter’s dynamic response. Additionally, a novel control strategy using feedforward and feedback mechanisms with Single Phase Shift (SPS) modulation is proposed to ensure seamless transitions between CC and CV modes, mitigating undershoot and overshoot issues. The research then presents an innovative control method that integrates droop-based techniques with a Linear Active Disturbance Rejection Control (LADRC) system, applied to an SPS-modulated DAB converter. This method enhances robustness against voltage fluctuations and load variations, stabilizing the charging process. It also reduces peak input current, enabling the use of lower-rated components, thus lowering the cost and complexity of EV charging stations. Advanced modulation techniques, specifically Dual-Phase Shift (DPS) modulation, are explored to optimize soft-switching capabilities. A three-loop LADRC control system with decoupled dq vector control of inductor current is proposed to optimize phase shift angles and reduce peak input current, further reducing system costs. Additionally, the research also compares modular DAB topologies under various metrics such as efficiency, system losses, and ripple characteristics. Results show that modular input-parallel output-parallel DAB converters offer superior efficiency, particularly under variable loads, reducing stress on power switches and allowing for cost-effective components. To manage power balance and reduce sensor use, an enhanced extended state observer (ESO) is proposed, improving system performance. Moreover, an adaptive module-shedding technique is proposed which dynamically adjusts active DAB modules based on real-time power demand, improving efficiency across a wide range of operating conditions. The innovations presented in this research address the multifaceted challenges of fast EV charging, focusing on efficient power conversion, system robustness, and cost-effectiveness, contributing to the global transition to sustainable transportation solutions.