Several previous research initiatives have highlighted the role of Information and Communication Technologies (ICT) as key enablers for decreasing energy usage in buildings. However, few advances have been achieved in underground public spaces. This paper introduces a novel intelligent energy management system for underground stations. The system implements artificial intelligence solutions for autonomous building system control, based on advanced control algorithms that can learn from previous operations and situations. The robustness needed to operate in public spaces is achieved through a seamlessly integrated monitoring grid with self-diagnosis mechanisms. A middleware platform integrates existing devices, subsystems and newly deployed sensor-actuator networks. Results obtained during the implementation of the system in a prototype underground station showed potential yearly energy savings ranging between 74,336 and 87,339 kW h. The highest energy savings potential was found in the ventilation subsystem (30.6% +/- 2.0%), followed by the lighting system (24.1% +/- 1.9%) and escalators (8.5% +/- 1.9%).

Peer Reviewed