MUMBAI, India, Aug. 12 -- Intellectual Property India has published a patent application (202641092476 A) filed by St. Josephs College Of Engineering; Mythiri Janarthanan; Kanishka S; and Mr. R. Sreekanth on July 30, 2026, for Machine Learning Based Solar Power Forecasting System.
Inventors include St. Josephs College Of Engineering; Mythiri Janarthanan; Kanishka S; and Mr. R. Sreekanth.
The application for the patent was published on August 07, 2026, under issue no. 32/2026.
Abstract: In the current scenario, solar photovoltaic systems have been one of the most significant renewable energy sources, but their generated electrical energy fluctuates depending on the weather conditions, which brings some problems for grid operations. Therefore, accurate forecasting of solar energy generation is a necessity to maintain stability in power grids and manage renewable energy. This paper proposes a machine learning algorithm for forecasting the power generation of solar PV panels based on several environmental and solar positioning indicators like temperature, humidity, cloud cover, radiation from the sun, and angle of the sun. Three models, which are linear regression, decision tree regression, and gradient boosting regression, were trained using the data. Preprocessing techniques, like dealing with missing data, normalization, and split train and test datasets, were utilized to enhance the performance of the models. The accuracy of all models was measured by the coefficient of determination (R²), Root Mean Squared Error (RMSE), and Mean Absolute Error (MAE). It turned out that the best model among all three is Linear Regression with R² = 0.86.
Disclaimer: Curated by HT Syndication.
