Electricity Demand Forecasting In Kerala Using Machine Learning Models

Abstract

In the contemporary era, energy has evolved into a necessity for day-to-day life. Forecasting the demand for electricity is an extremely challenging undertaking. Forecasting the electricity accurately helps to run the power systems efficiently and effectively.In this work, we have developed an electricity demand forecasting model that predicts electricity demand using our own dataset. Our dataset contains 10 years of data (January 2013 to December 2022).The electricity demand data was taken from the Kerala State Electricity Board (KSEB). Linear Regression (LR), Decision Tree (DT), Random Forest (RF), Support Vector Regression (SVR), K-Nearest Neighbors (KNN), XGBoost, Artificial Neural Network (ANN) based on Machine Learning (ML) methods have been applied to observe how these algorithms perform in forecasting electricity. The performance of the discussed ML methods has been analysed based on various evaluation metrics such as accuracy, MAE, MSE, and RMSE. The outcomes obtained from the analysis show that the Random Forest ML approach outperforms the other conventional ML approaches in terms of accuracy, MAE, MSE, and RMSE since it has the highest accuracy (82.72%) and the lowest MAE, MSE, and RMSE score (0.038, 0.002, and 0.054, respectively) compared to the other discussed conventional ML approaches.