Design and optimization of dual active bridge converter for Type-2 charging infrastructures using metaheuristic methods

Abstract

This study presents a pioneering approach in the design and optimization of Dual Active Bridge (DAB) converters for Type-2 Electric Vehicle (EV) charging infrastructure, utilizing advanced metaheuristic algorithms, specifically the Particle Swarm Optimization (PSO) method. It involves the application of PSO, a robust optimization technique inspired by social behavior patterns observed in nature, such as bird flocking or fish schooling. This method is adept at navigating complex, multi-dimensional solution spaces to identify optimal solutions that traditional methods might overlook. The focus of the study is the prevalent Type-2 charging stations, typically operating around the 7kW range. Given their widespread usage and cost-effectiveness, the research concentrates on optimizing the DAB converter for these stations. A gap in the literature regarding the dynamic analysis of component losses in DAB converters is identified and addressed through this study. The integration of PSO in this context is particularly innovative, given the unique challenges posed by Dual Active Bridge (DAB) converters. These challenges encompass the management of dynamic loads, varying power requirements, and the need to ensure efficient power transfer while maintaining system stability and reliability. The PSO algorithm tackles these challenges by simulating a group of particles (potential solutions) moving through the solution space. Each particle adjusts its trajectory based on its own experience and the collective wisdom of the swarm, effectively balancing exploration and exploitation strategies to converge on an optimal solution. In the component optimization phase, the PSO method is utilized to its full potential. This approach is pivotal for tuning components such as transformers, inductors, MOSFETs as switching elements, and capacitors. PSO's strength in efficiently exploring complex, multi-dimensional solution spaces allows for the optimal selection and configuration of these components, considering factors such as efficiency, loss, cost, and size. Contrastingly, the system optimization process diverges from the use of PSO. Instead, it employs a comprehensive frequency-sweeping methodology. This technique involves systematically varying the switching frequency of the DAB converter and evaluating the performance at each frequency. The objective is to identify the optimal switching frequency that minimizes losses and maximizes efficiency, thereby ensuring the converter's compatibility with a broad range of operational scenarios. The study also includes extensive simulation and validation work using tools like Ansys Maxwell, PSIM, and Simulink. These simulations serve to verify the theoretical models and provide practical insights into the real-world applicability of the optimized converter. Additionally, the study conducts a thorough thermal analysis of switching elements and assesses the leakage inductance of transformers, integrating all losses into the simulation environment for a realistic evaluation. In conclusion, this study significantly contributes to the field of EV charging infrastructure by presenting a combined metaheuristic and frequency-sweeping-based optimization framework for DAB converters. This approach yields a more efficient, reliable, and cost-effective solution for Type-2 EV charging stations, heralding advancements in smart charging technologies and sustainable transportation systems.