FPGA-Based Model Predictive Control for Three-Phase Smart Grid Automation

This study presents a smart real-time control framework for power management in virtualized grid environments for emerging African smart energy systems. The system coordinates power flow among distributed energy resources, including conventional generators and virtual microgrids, to ensure stable and reliable grid operation under unstable supply and fluctuating demand. An integer-order compartmental model is used to describe system dynamics under disturbances and varying operating conditions. A Model Predictive Control (MPC) scheme is implemented on a Field-Programmable Gate Array (FPGA) to achieve high-speed computation and low-latency control. A fuzzy logic supervisory layer improves adaptability under uncertainty and nonlinear conditions. Numerical simulations based on the Runge-Kutta method show stable, bounded system behaviour with low error propagation under optimal parameter settings. ( β i = 0.0003, σ = 5 × 10 −5 , ν = 0.03). Sensitivity analysis identifies the fuzzy adaptive gain ν as a key stability parameter. Results demonstrate improved transient response, enhanced dynamic stability, and robust performance. The framework demonstrates strong theoretical potential for resilient, intelligent smart grid automation across Africa.