Real-time Machine Learning-Based Physical Model for Estimating Photovoltaic Output Power Using Meteorological Data

This project focuses on creating a machine learning-based model to estimate photovoltaic (PV) output power using meteorological data. Solar energy is heavily influenced by environmental factors such as temperature, humidity, wind speed, and cloud coverage. Accurate predictions of PV output are crucial for optimizing the utilization of solar energy systems. The model employs machine learning algorithms including LGBMRegressor, AdaBoostRegressor, RandomForestRegressor, and Feedforward Neural Network (FNN) to forecast the amount of power generated by PV systems. By analyzing historical data that includes various meteorological factors, the model is trained to estimate the energy produced by PV panels based on specific weather conditions. The project uses a Flask-based web application to allow users to input location and weather data and receive output power predictions. The ultimate goal is to provide a reliable and efficient method for estimating PV output, which can assist in improving energy management, helping solar energy producers optimize their systems and enhancing grid stability.

Keywords: photovoltaic, power estimation, meteorological data, machine learning, LGBMRegressor, AdaBoostRegressor, RandomForestRegressor, Feedforward Neural Network, weather conditions, energy optimization