Predicting increased electricity consumption during severe weather conditions using Artificial Intelligence
The unprecedented levels of climatic changes has led to increased unpredictability of the users’ electricity consumption, causing Energy Management Systems (EMS) to make an unplanned shutdowns of electricity grids. Accurate energy/electricity consumption prediction is an essential component in ensuring reliability of the grid and providing steady electricity output during severe weather events. Machine learning (ML) and Deep Learning (DL) methods is recognized as one of the suited approach for understanding co-relations between weather conditions and electricity consumption. However, there has not been the avalanche of the usage of ML methods for accurately predicting increased electricity consumption using weather as an input data. Our project focuses on using the ML, DL methods to predict the increased electricity consumption from the baseline consumption during extreme weather conditions.
We used different machine learning techniques for this project: Linear Regression, Logisitic Regression, Support Vector Classifier and Deep Neural Network.