Citation
Mehdi Maasoumy, Meysam Razmara, Mahdi Shahbakhti, Alberto Sangiovanni-Vincentelli. "Handling model uncertainty in model predictive control for energy efficient buildings". Energy and Buildings, 77:377-392, July 2014; http://www.sciencedirect.com/science/article/pii/S0378778814002771.

Abstract
Model uncertainty is a significant challenge to more widespread use of model predictive controllers (MPC) for optimizing building energy consumption. This paper presents two methodologies to handle model uncertainty for building MPC. First, we propose a modeling framework for online estimation of states and unknown parameters leading to a parameter-adaptive building (PAB) model. Second, we propose a robust model predictive control (RMPC) formulation to make a building controller robust to model uncertainties. The results from these two approaches are compared with those from a nominal MPC and a common building rule based control (RBC). The results are then used to develop a methodology for selecting a controller type (i.e. RMPC, MPC, or RBC) as a function of building model uncertainty. RMPC is found to be the superior controller for the cases with an intermediate level of model uncertainty (30-67%), while the nominal MPC is preferred for the cases with a low level of model uncertainty (0-30%). Further, a common RBC outperforms MPC or RMPC if the model uncertainty goes beyond a certain threshold (e.g. 67%).

Electronic downloads

• http://www.sciencedirect.com/science/article/pii/S0378778814002771
• https://www.src.org/library/publication/p070187/

• HTML

  Mehdi Maasoumy, Meysam Razmara, Mahdi Shahbakhti, Alberto
  Sangiovanni-Vincentelli. <a
  href="http://www.terraswarm.org/pubs/310.html"
  >Handling model uncertainty in model predictive control
  for energy efficient buildings</a>, <i>Energy
  and Buildings</i>, 77:377-392, July 2014;
  http://www.sciencedirect.com/science/article/pii/S0378778814002771.

• Plain text

  Mehdi Maasoumy, Meysam Razmara, Mahdi Shahbakhti, Alberto
  Sangiovanni-Vincentelli. "Handling model uncertainty in
  model predictive control for energy efficient
  buildings". <i>Energy and Buildings</i>,
  77:377-392, July 2014;
  http://www.sciencedirect.com/science/article/pii/S0378778814002771.

• BibTeX

  @article{MaasoumyRazmaraShahbakhtiSangiovanniVincentelli14_HandlingModelUncertaintyInModelPredictiveControlForEnergy,
      author = {Mehdi Maasoumy and Meysam Razmara and Mahdi
                Shahbakhti and Alberto Sangiovanni-Vincentelli},
      title = {Handling model uncertainty in model predictive
                control for energy efficient buildings},
      journal = {Energy and Buildings},
      volume = {77},
      pages = {377-392},
      month = {July},
      year = {2014},
      note = {http://www.sciencedirect.com/science/article/pii/S0378778814002771},
      abstract = {Model uncertainty is a significant challenge to
                more widespread use of model predictive
                controllers (MPC) for optimizing building energy
                consumption. This paper presents two methodologies
                to handle model uncertainty for building MPC.
                First, we propose a modeling framework for online
                estimation of states and unknown parameters
                leading to a parameter-adaptive building (PAB)
                model. Second, we propose a robust model
                predictive control (RMPC) formulation to make a
                building controller robust to model uncertainties.
                The results from these two approaches are compared
                with those from a nominal MPC and a common
                building rule based control (RBC). The results are
                then used to develop a methodology for selecting a
                controller type (i.e. RMPC, MPC, or RBC) as a
                function of building model uncertainty. RMPC is
                found to be the superior controller for the cases
                with an intermediate level of model uncertainty
                (30-67%), while the nominal MPC is preferred for
                the cases with a low level of model uncertainty
                (0-30%). Further, a common RBC outperforms MPC or
                RMPC if the model uncertainty goes beyond a
                certain threshold (e.g. 67%).},
      URL = {http://terraswarm.org/pubs/310.html}
  }
  

Posted by Mehdi Maasoumy on 10 May 2014.