Electricity Consumption Analysis and Prediction

Abstract

In our rapidly evolving world, the demand for electricity has surged due to its vital role in driving national development across various sectors. Accurate forecasting of electricity demand is crucial for effective energy resource planning and management. The Electricity Consumption Analysis and Prediction System utilized a comprehensive dataset and employed KNN, linear regression, and XGBoost algorithms to predict electricity consumption. The data for our project has been collected from the Grid Office in Baneshwor. After meticulous data cleaning, feature engineering, and integration, XGBoost emerged as the optimal model, showcasing superior accuracy with Mean Squared Error (MSE) of 0.03, Root Mean Squared Error (RMSE) of 0.18, and an impressive R2-score of 0.95. The system effectively visualizes analyzed data and model predictions using charts and graphs, providing users with intuitive insights. The system generates predictions for electricity consumption on a monthly, daily, and yearly basis for the Baneshwor Area.