Prediction of Solar Power Generation and Ground Area Estimation Using KNN Regression

Abstract

This paper focuses on making use of historical weather data to predict the potential power generation through the application of the k-Nearest Neighbours (k-NN) algorithm and subsequent estimation of the precise area needed to accommodate the solar panels. Effect of variables such as ambient temperature, solar irradiation, solar panel temperature, latitude and longitude, topology, and cloud coverage is studied on the solar power generation capacity. The model is trained to forecast the power output of solar panels based on three environmental variables of ambient temperature, solar irradiation and solar panel temperature. The k-NN algorithm, known for its simplicity and efficiency, is employed to capture the patterns inherent in historical datasets. This methodology allows for the accurate prediction of solar power generation under diverse weather conditions. The incorporation of temperature data, both ambient and solar panel-specific, enhances the precision of the model. The research outcomes not only contribute to the advancement of solar energy forecasting but also help optimize the energy grid management using the k-NN methodology.