Predictive Networked Building Control


Buildings are responsible, directly and indirectly, for 40% of US carbon dioxide emissions. Reductions of 70% in energy use in buildings are required to achieve emissions reduction goals for the building sector set by a number of organizations, including the California Public Utilities Commission. Achieving this goal will require the use of high efficiency heating and cooling systems, which are more challenging to control than conventional systems.

Energy-optimal operation of such a system requires solution of a complex distributed and constrained control problem. Hard constraints arise from maximal emissions specifications on CO2, energy consumption and human comfort. Orchestrating hundreds of actuators by using predictive knowledge of heat transfer dynamical models, weather forecasts and people schedule is a very complex control tasks which easily becomes intractable if approached with a centralized design methodology.

Research Overview

Our goal is to use the constrained distributed control theory to significantly improve the performance of these systems. Improved performance would include reduced energy consumption, reduced peak demand and tighter regulation for thermal comfort. In the proposed mathematical contest, the communication and sensing graph is a design variable which can be changed in real time and add additional degrees of freedom to the system: rooms sensors and actuators in general might “talk to” their spatial neighbors but there might exist better communication policies. Also, cameras and presence detectors can provide to neighboring nodes of the network a preview on people motion across the building. Finally, the “intent” of each actuator can be communicated and transmitted to other node and help them make better predictions.

Current Research Projects

  • 2011-2013MPC for EPMO (Energy Performance Monitoring and Optimization),  jointly with United Technologies Research Center (UTRC). 
This research focuses on the control of HVAC system of two buildings and a chiller plant for demonstration of a campus-scale EPMO prototype on a DOD campus.  
The objective of this research is to design a predictive controller in order  to minimize energy consumption of the two buildings with a large number of zones and the associated chiller plant in real time.
  • 2013-2018. Model Predictive Control for low energy HVAC systems: theory and tools  (jointly with Lawrence Berkeley National Laboratories)
The objective is to collaborate with LBNL to develop Model Predictive Control (MPC) techniques and integrate them with EnergyPlus . The focus will be on low energy HVAC systems and the impact of uncertain disturbances in the control design process and in the choice of optimal control actions.
  • 2013-2015. Model Predictive Control for chiller plants  (jointly with Honeywell)

Past Projects

  • 2008-2009: MPC For UC Merced Campus (jointly with  US Department Of Energy (DOE), Lawrence Berkeley National Laboratory (LBNL), United Technologies Research Center (UTRC) and University of California at Merced )

This research focuses on the modeling and the control of the thermal energy storage on the campus of the University of California, Merced, USA. The campus has been designed to be a ''living laboratory'' and has a significantly enhanced level of instrumentation in order to support the development and demonstration of energy-efficient technologies and practices.It consists of a chiller plant (three chillers redundantly configured as two in series, one backup in parallel), an array of cooling towers, a 7000 m^3 chilled water tank, a primary distribution system and secondary distribution loops serving each building of the campus. The two series chillers are operated each night to recharge the storage tank which meets campus cooling demand the following day.
The objective of this research is to design a predictive controller in order to minimize energy consumption while satisfying the unknown but bounded cooling demand of the campus buildings and operational constraints.
  • 2009-2010:Wireless Platform for Energy-Efficient Building Control Retrofits
For the foreseeable future the largest opportunity to reduce DOD facility energy consumption will come from retrofits and renovations of existing buildings.   One promising technology is whole-building optimal control which has the potential to reduce building energy consumption by 3-10% (0.5-1.7 of the 17 quads of energy consumed by US commercial buildings)[i], and reduce peak electrical demand by 10-20%, an important metric in localities where power distribution is constrained. 
There are two objectives of this research: 
1) To demonstrate the energy efficiency gains achievable in small to medium sized buildings with Model Predictive Control (MPC), a form of whole-building optimal control.
2) To demonstrate the reduction in first costs achievable with a Wireless Sensor Network (WSN) – based building HVAC control system compared to a conventional wired system.