Cluster-based energy consumption forecasting in smart grids

Abstract

Accurate energy saving estimates are important for development of energy efficient projects and for demonstrating their cost-effectiveness. Increasingly, commercial buildings have an advanced measurement infrastructure, which has led to the availability of high-frequency interval data. This data can be used in a number of energy efficiency tasks, including anomaly detection, control and optimization of heating, ventilation, and air cooling systems. These data sets enable the application of advanced statistical training models and therefore lead to accurate estimates of energy savings. The K-means algorithm is a powerful tool that has been used in areas with intensive data mining applications, including ecology, computer vision, biology. In this paper, we propose a method for modeling the basic energy profile based on the clustering of consumers. A recently published test procedure was used to assess the accuracy of this method, and the estimate was based on analysis of data from 288 commercial buildings. The training periods of the model varied, and some predictive accuracy indicators were used to estimate the accuracy of the model. The results showed that the use of our model improved the determination and cross-validation parameters in more than 80 percent of cases in comparison with the best model used in the industry, which is based on regression, and in comparison with the random forest, decision trees.