Predictive Control Strategies for CC-CV Charging of Offboard Chargers: Formulation, Comprehensive Analysis, and Experimental Validation

Abstract

When conventional voltage-oriented control (VOC) and direct power control (DPC) strategies are applied to electric vehicle (EV) chargers, they can compromise their performance owing to a cascaded control structure with limited bandwidth and controller-gain sensitivity under varying charging conditions. Alternatively, finite control set model predictive control (FCS-MPC) has emerged as a promising solution due to its direct use of discrete switching states and multiobjective optimization capabilities. However, a systematic investigation and comprehensive analysis of current-based and power-based FCS-MPC strategies for offboard EV chargers in terms of grid power quality, charger efficiency, battery side ripple, computational complexity, and control robustness under varied charging profiles has not been explored yet. To fill this research gap, this article presents a comprehensive formulation, implementation, validation, and analysis of predictive current control (PCC) and predictive power control (PPC) strategies for a two-stage offboard charger operating under constant current-constant voltage (CC-CV) charging conditions. In addition, a new charge cycle efficiency metric is introduced to quantify overall charger efficiency during the complete charging process. To validate the effectiveness of both methods experimentally, a two-stage offboard charging system with a dSPACE MicroLabBox control platform is developed to charge a 2.4 kWh, 400 V battery pack. The charger’s performance with the PCC and PPC schemes is holistically evaluated under steady-state, transient, CC-CV charging, and parameter mismatch conditions. Finally, the tradeoffs and selective guidelines of both predictive control strategies for EV charging applications are highlighted.