A Scalable and Adaptive Model-Predictive Distributed Control of Building HVAC Systems
This technology offers a novel model-predictive distributed control method and a flexible and cost effective IoT implementation architecture for energy saving in building HVAC systems. It is scalable and real-time optimally responsive to changes in a large building that has more than 500 zones via a patented token-based HVAC scheduling strategy. The larger the building, the higher the energy saving potential, due to its novel coordinated HVAC scheduling approach. It is autonomously adaptive to the building operational environment via effective system identification techniques, including machine learning techniques, on real-time data attainable from the proposed IoT infrastructure. It is also occupant-centric, i.e., capable of learning and addressing individual human comfort requirements. The technology is applicable to any new or old VAV (or VRV) HVAC system without any need of major retrofitting on existing HVAC controllers and data acquisition systems, due to its highly flexible plug-and-play implementation architecture. It can also be used to convert an old HVAC system into a highly automated and intelligent one, allowing a building owner to check remotely, via commonly used mobile devices, the status of the building HVAC system and initiate supervisory control for better performance, thus, can enhance the effectiveness of existing BMS and building automation. The technology provider is seeking for industry partners to commercialise the technology.