<![CDATA[The development of renewable energy technology has encouraged the utilization of photovoltaic (PV) systems as an alternative energy source. However, variations in irradiance and temperature cause changes in PV output power, making it difficult to maintain the Maximum Power Point (MPP). This study aims to design and analyze a Maximum Power Point Tracking (MPPT) system based on an Artificial Neural Network (ANN) integrated with a buck-boost converter for a PV system. The system was simulated using MATLAB/Simulink with a 100 Wp PV module and a Feedforward Neural Network (FNN) trained using the Levenberg–Marquardt (LM) algorithm. The FNN uses irradiance and temperature as inputs and generates the duty cycle as the converter control output. Testing was conducted under 16 conditions with irradiance levels of 300–1000 W/m² and temperatures of 25–40°C. The training results achieved a correlation coefficient of 0.92, with a duty cycle ranging from 0.40 to 0.45. The highest PV power was 99.84 W at 1000 W/m² and 25°C, with an output power of 92.85 W and an efficiency of 93.00%. The results demonstrate that the FNN-based MPPT system can adjust the duty cycle according to changes in PV conditions.]]>
Copyrights © 2026
