Published Paper Details:

The global transition to sustainable energy necessitates significant improvements in the efficiency of renewable sources like solar power. Traditional methods for optimizing photovoltaic (PV) panel performance often rely on static positioning or simple sun-tracking, failing to account for complex, real-time environmental variables. This paper explores the application of a Physics-Informed Neural Network (PINN) to dynamically maximize the energy output of a PV system. By integrating the fundamental physical principles of photovoltaics (the single-diode model) with a machine learning framework that processes real-time weather data (irradiance, temperature, cloud cover), the proposed system predicts the optimal tilt and orientation angles for a PV panel. A simulated case study demonstrates that the PINN model increases daily energy capture by approximately 18.5% compared to a fixed-angle system and by 7.2% over a conventional dual-axis tracker, by more intelligently responding to diffuse irradiance and cloud-transition periods. This work underscores the potent synergy between physics-based modeling and artificial intelligence in addressing critical challenges in sustainable energy, a key pillar for societal growth and achieving Sustainable Development Goals (SDGs).