Date of Award


Degree Type


Degree Name

Master of Science (MS)


Electrical and Computer Engineering

Committee Chair(s)

Rose Qingyang Hu


Rose Qingyang Hu


Tung Thanh Nguyen


Don Cripps


Smart Grids are the next generation electrical grid system that utilizes smart meter-ing devices and sensors to manage the grid operations. Grid management includes the prediction of load and and classification of the load patterns and consumer usage behav-iors. These predictions can be performed using machine learning methods which are often supervised. Supervised machine learning signifies that the algorithm trains the model to efficiently predict decisions based on the previously available data.

Smart grids are employed with numerous smart meters that send user statistics to a central server. The data can be accumulated and processed using data mining and machine learning techniques to extract meaningful insights. Forecasting of future grid load (electricity usage) is an important task for gaining intelligence in the gird. Accurate forecasting will enable a utility provider to plan the resources and also to take controlled actions to balance the supply and the demand of electricity. This forecasting can be achieved using machine learning based predictive models.

In this project, a predictive system is designed that uses data mining and machine learning techniques to process the smart meter data and to use it as training data for the model. The main objective of this project is to forecast short term to mid-term load for the grid entity. The outcomes are backed with visualizations to make the data and results more user readable.