September 29, 2022
Staff Accomplishment

Advancing Model Predictive Control for Buildings

Researchers contribute to most-cited paper

PNNL's Jan Drgona and Draguna Vrabie

PNNL’s Ján Drgoňa (left), lead author, and Draguna Vrabie have attained a most-cited article in Annual Reviews in Control.

(Photos by Andrea Starr, composite image by Shannon Colson | Pacific Northwest National Laboratory)

Two researchers at Pacific Northwest National Laboratory (PNNL) are part of an international team that authored a most-cited paper articulating strategies for implementing Model Predictive Control (MPC) in commercial and other buildings.

PNNL’s Ján Drgoňa, a data scientist and the paper’s lead author, partnered with 10 other experts, including his PNNL colleague Draguna Vrabie. The team represented research organizations in Belgium, Denmark, Slovakia, and the United States.

The paper, “All You Need to Know About Model Predictive Control for Buildings,” was published in 2020 in Annual Reviews in Control. The journal’s current most-cited papers reflect those published since 2019.

Buildings as climate benefactors

The paper notes that buildings worldwide collectively consume significant amounts of energy. As climate concerns take center stage, advanced control methodologies, compared to existing techniques, offer the potential to better manage the operation of building systems, such as heating, ventilation, and air-conditioning units, reducing energy consumption and greenhouse gas emissions. Advanced methods can also improve occupant comfort and perhaps even enable direct interactions between building systems and the electrical grid, which help balance electricity demand and incorporate more clean energy into the power system.

MPC has shown considerable promise as an advanced control approach for building operations, particularly heating and cooling. As the paper explains, the method employs a mathematical model of a building to predict the building’s future behavior. By using the predictions, MPC can optimally choose control actions based on a given objective while taking into account occupant comfort, technological constraints, and weather forecasts in a systematic and flexible way.

One of the paper’s key objectives is to help make MPC concepts more accessible to a broad range of researchers and practitioners with different engineering backgrounds.

“Widespread use of MPC in new buildings is likely a decade away,” Drgoňa said. “This paper hopefully moves us a step closer to broader deployment. It provides a framework for MPC applications for energy-efficient building controls. The paper also includes a comprehensive overview of different modeling approaches, problem formulations, solution methods, software tools, and performance assessments focusing on real-world applications.”

A multi-faceted research effort

The team’s work emerged from a project conducted under the International Building Performance Simulation Association and received funding support from the Department of Energy’s Building Technologies Office, the European Union, and the VITO research organization.

In addition to Drgoňa and Vrabie, who work in PNNL’s Physical and Computational Sciences Directorate, the team included Javier Arroyo, Iago Cupeiro Figueroa, Enric Perarnau Ollé, and Lieve Helsen, representing KU Leuven in Belgium; David Blum, Donghun Kim, and Michael Wetter representing Lawrence Berkeley National Laboratory; Krzysztof Arendt of the University of Southern Denmark; and Juraj Oravec of the Slovak University of Technology.