Abstract:
This thesis presents an investigation of the Maximum Power Point Tracking (MPPT) approach for solar power systems utilizing a DC-DC flyback converter operated by a Neural Network (NN) for a gridconnected PV system. MATLAB/ Simulink is used to model and simulate the entire system, then the NN framework is used to train the NN. The outcomes of the proposed MPPT controller are then compared to a perturb and observe (P&O) MPPT algorithm. The parameters Vpv, Ipv, Ppv, and voltage of the DC link of the PV array also inverter voltage and inverter current are taken into account in this thesis research. The modeling and simulation of the indicated MPPT approaches are simulated on a 10.457-kWP PV array that forms as a grid-connected powers generator. According to the findings, the recommended NN-based MPPT controller performs better and produces good results in terms of Maximum Power Point (MPP) under a variety of weather conditions. It was also discovered that intelligent-based MPPT algorithm has lower power rippling, high-speed response, harmonics reduction, and quick outcomes than traditional P&O algorithms.
