Predicting Future Hourly Residential Electrical Consumption: A Machine Learning Case Study

Traditional whole building energy modeling suffers from several factors, including the large number of inputs required to characterize the building, the specificity required to accurately model building materials and components, simplifying assumptions made by underlying simulation algorithms, and the gap between the as-designed and as-built building. Prior work has attempted to mitigate these problems by using sensor-based machine learning approaches to statistically model energy consumption. We refer to this approach as sensor-based energy modeling (sBEM). However, a majority of the prior sBEM work focuses only on commercial buildings. The sBEM work that focuses on modeling residential buildings primarily focuses on monthly electrical consumption, while commercial sensor-based models focus on hourly consumption. This means there is not a clear indicator of which machine learning approach best predicts next hour residential consumption, since these methods are only evaluated using low-resolution data. We address this issue by testing seven different machine learning algorithms on a unique residential data set, which contains 140 different sensors measurements, collected every 15 minutes. In addition, we validate each learner’s correctness on the ASHRAE Great Energy Prediction Shootout, using the original competition metrics. Our validation results confirm existing conclusions that Neural Network-based methods perform best on commercial buildings. However, the results from testing on our residential data set show that Feed Forward Neural Networks (FFNN), Support Vector Regression (SVR), and Linear Regression methods perform poorly, and that Least Squares Support Vector Preprint submitted to Buildings and Energy October 28, 2011 Machines (LS-SVM) perform best – a technique not previously applied to this domain.