<![CDATA[General Background : The rising global penetration of photovoltaic (PV) systems necessitates inverter technologies capable of maintaining power quality under dynamic conditions. Specific Background : Finite Set Model Predictive Control (FS-MPC) offers fast dynamic response but remains limited by fixed, manually tuned weighting factors that do not adjust to changing irradiance, loads, or grid disturbances. Knowledge Gap: Existing studies rarely provide real-time adaptive tuning mechanisms that remain computationally feasible for embedded inverter applications. Aim: This study proposes a hybrid control strategy that integrates FS-MPC with Particle Swarm Optimization (PSO) to automatically adjust cost-function weights during operation. Results: Simulations in MATLAB/Simulink show that the PSO-FS-MPC controller achieves lower THD (2.1%), faster settling time (8 ms), and higher conversion efficiency (96.8%) than conventional FS-MPC and PI-SPWM methods across irradiance drops, load variations, grid sags, and partial shading. Novelty: The method performs continuous, real-time parameter adaptation using a lightweight optimization layer without exceeding computational limits. Implications: These findings indicate that adaptive predictive control can support more stable, efficient, and grid-compliant PV inverter operation under realistic dynamic scenarios. Highlights: Introduces real-time adaptive tuning for FS-MPC using PSO. Achieves superior THD, efficiency, and transient response under dynamic conditions. Demonstrates computational feasibility for practical inverter implementation. Keywords: PV Inverter, Model Predictive Control, Particle Swarm Optimization, Power Quality, Adaptive Tuning]]>
