Electrical Load Demand Forecast for Gujarat State of India using Machine Learning Models

The work is concentrates on forecasting electrical loads of Gujarat state of India. Day ahead load demand forecasting is proposed by utilizing historical load demand data from year 2022 to 2023. The dataset, acquired for these years, are processed using Python scripts to filter and organize the information with a 15-minute timestamp granularity. Subsequently, Regression based machine learning algorithms are applied on Load Demand data of Gujarat to forecast day ahead load demand. The proposed work involves a comparative analysis of the performance of Linear Regression, Support Vector Regression and Regression Tree algorithms with their classes models, to enhance accuracy and efficiency in load forecasting. The results are evaluated based on metrics such as Mean Absolute Error (MAE) and Mean Absolute Percentage Error (MAPE) The Robust Linear Regression model outperforms all the classes with MAE 70.252 MW of and MAPE of 0.176%. This research aims to contribute valuable insights for making decision in the power sector by offering a refined and effective methodology for load prediction, potentially saving significant resources for utility companies.