Argonne National Laboratory Renewable Energy Research & Development U.S. Department of Energy
           
 Search

Building Efficiency:
Building-wide, Proactive Energy Management Systems for High-Performance Buildings

machine learning

Achieving and sustaining energy savings depends on advances in energy management (EM)--for example, the ability to exploit occupancy, energy prices, and weather trends when optimizing building system performance. Reducing the computational complexity of building-wide EM systems will make them more desirable.

Through a CRADA with BuildingIQ, Argonne is developing a proactive EM system with occupancy-based control to address these challenges. While building operations are inherently dynamic, with changes occurring quickly in external conditions, building systems respond slowly to these changes and provide an opportunity to optimize energy use, occupant comfort, and system responsiveness. To capture that opportunity, Argonne's system will use adaptive building-wide predictive models that account for zone topology, energy and mass balances, HVAC systems, real-time occupancy, energy prices, and weather forecast information.

EM systems rely on iterative solutions of complex optimization problems, which need to be solved quickly and reliably to ensure appropriate real-time performance. Fast optimization algorithms also enable implementation with inexpensive, commodity hardware, thereby enabling widespread deployment. The research team is leveraging Argonne's expertise in numerical optimization to deploy state-of-the-art and open-source optimization functionality and developing model reformulations and warm-starting strategies. These capabilities will enable set-point updates in a few minutes for large and highly detailed building models and long forecast horizons, enhancing system responsiveness and robustness.

The algorithms are being implemented in BuildingIQ's system, which uses building models constructed automatically from sensor data and machine learning techniques. This approach avoids the need for expensive model development tasks and simulation engines that limit deployment.

A proactive EM system is currently in use at the Advanced Photon Source office building on the Argonne campus. Argonne and BuildingIQ are demonstrating novel occupancy-based controls to achieve increased energy savings.

This project is being funded by U.S. Department of Energy's Building Technologies Program.

More

August 2012

CONTACT

Victor Zavala
vzavala@mcsanl.gov

 

U.S. Department of Energy Office of Science | UChicago Argonne LLC
Privacy & Security Notice | Contact Us | Site Map | Search