This paper presents a novel hybrid approach based on fuzzy logic controller and gravity search algorithm to track the global maximum power point of a network-connected photovoltaic system in partially shaded conditions. One of the most critical issues in this context, which has been neglected in previous studies, is the consideration of speed and accuracy at the same time. Hence, this paper uses a boost converter with a fuzzy logic controller to increase the model's accuracy. Also, the speed of the method is increased by utilizing the gravity search algorithm. Finally, maximum power is subjected to a power network via a three-phase multi-level inverter. Simulation results show the proposed method's performance and accuracy in tracking the PV system's maximum power point with high-speed responses in partial and variable shadow conditions.