Multi-Source Energy Management in A Grid-Synchronized Bi-directional EV Charging Architecture with Error-Modified Adaptive Control

This paper presents a robust error-modified (E-Mod) adaptive tracking control algorithm designed for fundamental grid frequency estimation and multi-resonant direct power control, which regulates bi-directional EV charging operations under distorted grid voltage conditions. The non-linear phase and frequency detection algorithm self-updates with each analog signal sample, ensuring rapid reference command tracking even in the presence of grid non-idealities. The algorithm's robust and stable controller enhances the charging profile by reducing DC offset error, switching noise, harmonic distortion, and sensor measurement error. The resonant control architecture incorporates both phase and frequency information at dominant harmonic multiples, enabling regulated EV charging with unity power factor (UPF) in resilient grid conditions. A linear state model with current-regulated phase-shift operation facilitates DC-DC power conversion for variable battery charging. The phase-shift operation in the second stage is updated by equivalent voltage feed-forward compensation to improve EV charging dynamics. The charging architecture and related control implementation are designed, simulated, and validated using a laboratory prototype of a 3.3kW Level-I bi-directional EV charger.