Optimization approach of LLC resonant converter parameters for electric vehicles based on interior-point algorithm

Abstract

This paper presents an optimization strategy for the design of an isolated bidirectional LLC resonant DC–DC converter used in electric vehicle chargers. By employing the interior-point algorithm, we optimize the resonant tank parameters ($C_r$, $L_r$ and $L_m$), expand the soft-switching region and reduce switching losses across wide battery voltage (240 V to 430 V) and power (1 kW to 11 kW). This approach expands the soft-switching region to improve efficiency in both charging and discharging modes, while simultaneously modeling and optimizing both Grid-to-Vehicle (G2V) and Vehicle-to-Grid (V2G) operation using analytically derived frequency equations. Additionally, it reduces the saturation zone by more than 50% and finds optimum values of resonant parameters $C_r$, $L_r$, and $L_m$. Unlike prior works that use heuristic optimization or treat G2V and V2G separately, this method combines both modes into a single constrained nonlinear optimization problem, offering a unified and generalized design approach. These enhancements lead to a more compact design, greater stability, and an optimal balance between efficiency and component size. Simulation results confirm the effectiveness of the method, with improved frequency controllability and converter efficiency across a wide range of operating points compared to existing reference designs.