This article proposes a finite set model predictive control (FS-MPC) strategy for a three-phase, two-stage photovoltaic (PV) and battery-based hybrid microgrid (HMG) system. The system incorporates parallel inverters with dual DC-link capacitors connected to a shared DC grid, enabling enhanced reliability and efficient power-sharing. A discrete-time HMG model is developed to predict key system parameters such as grid, circulating, and offset currents. To reduce computational complexity, the FS-MPC selectively employs 30 out of 64 switching vectors, ensuring faster processing without sacrificing performance. The system integrates an incremental conductance-based maximum power algorithm (IC-MPA) to achieve efficient PV energy extraction and a bidirectional converter model to regulate battery charging/discharging operations, maintaining DC-link voltage stability. A centralized energy management technique (CEMT) is also introduced to optimize energy flow and enhance system performance. The proposed approach is validated through comprehensive software simulations and hardware experiments, demonstrating significant improvements in power quality (PQ) and reliability (PR) under dynamic conditions. Key contributions include enhanced harmonic compensation, frequency instability mitigation, and faster response times, highlighting the practical effectiveness of the system in real-time hybrid microgrid applications.