Abstract: This study introduces an innovative control architecture for a three-port, 5 kW solar-integrated bidirectional electric vehicle charging platform, referred to as FCS-MPC-VSG-FO. A Finite Control Set Model Predictive Controller (FCS-MPC) is combined with a Virtual Synchronous Generator (VSG) emulation layer. This is made even stronger by a Fractional-Order Super-Twisting Sliding Mode Controller (FO-STSMC, α = 0.85) and a Multi-Variable Compensating (MVC) notch filter. The unified architecture fixes problems with poor transient recovery, grid-frequency susceptibility, and poor current quality when parameters are not known. The VSG layer acts like virtual inertia with Jvsg = 0.12 kg·m² and damping Dv = 18 N·m·s/rad. It keeps the frequency in check without any extra hardware. The PV boost stage is controlled by an Adaptive Variable-Step P&O (AVSP&O) algorithm that converges to the MPP in less than 15 ms and has an efficiency of 99.1%. Nine Lyapunov energy-balance requirements ensure stability in a closed loop. The findings from MATLAB/Simulink show that the grid current THD is less than 2.1%, the DC-bus variation is within ±1.4%, and the overall weighted efficiency is 96.8%. Comparative benchmarking against seven PI-based techniques shows that this one is better at transient responsiveness, harmonic rejection, and providing extra grid services.

Keywords: Adaptive MPPT, bidirectional EV charger, FCS-MPC, fractional-order super-twisting SMC, virtual synchronous generator, Lyapunov stability, harmonic filter, and photovoltaic integration.

Downloads: | DOI: 10.17148/IJIREEICE.2026.14541