Occupancy and occupant number detection for energy saving in smart buildings via machine learning techniques

Abstract

In this study, various machine learning techniques are applied for occupancy detection to provide occupancy-based energy savings in smart buildings. Occupancy detection can be achieved using environmental data obtained via various environmental sensors placed in smart environments. This study focuses on energy saving in smart buildings with occupancy detection, and avoiding unnecessary sensor use by determining which features are more effective in detecting occupancy by utilising a sample dataset. Sensor information considered as features and tested using various machine learning algorithms. In this context, both occupancy detection and occupant number detection classification are realised, and an exergy analysis is presented.