Register now After registration you will be able to apply for this opportunity online.
This opportunity is not published. No applications will be accepted.
Model Predictive Control for energy optimization in an occupied apartment
The aim of this semester project is to develop and implement a Model Predictive Control strategy based on linear state space models to minimize the energy consumption while maintaining the occupant thermal comfort of a unit in the NEST building at Empa in Dübendorf.
Keywords: model predictive control, building energy systems, system identification, MPC, optimization
Buildings are responsible for 40% of the primary energy consumption in Switzerland and for roughly one third in all other OECD countries. Improving the energy efficiency of buildings has therefore a significantly positive impact on the mitigation climate change.
Model predictive control (MPC) is a control strategy that has the potential to significantly reduce the energy consumption of most building energy systems. This is done by optimizing the control inputs (heating or cooling) in order to minimize a cost function over a receding horizon.
The project is part of a series of experiments with different predictive controllers in the apartment. The access to measurements and actuators is therefore already well established and the student can start their work with relatively little overhead.
If you are a motivated student and are interested in this project, please send an email with a short motivation letter, your transcript of records and CV. Experience with MPC, system identification, building thermal modelling, Python or Matlab are highly beneficial. You will be located at the Urban Energy Systems Laboratory at Empa, your supervising professor will be Prof. John Lygeros of the Automatic Control Laboratory at D-ITET ETH Zürich.
Buildings are responsible for 40% of the primary energy consumption in Switzerland and for roughly one third in all other OECD countries. Improving the energy efficiency of buildings has therefore a significantly positive impact on the mitigation climate change.
Model predictive control (MPC) is a control strategy that has the potential to significantly reduce the energy consumption of most building energy systems. This is done by optimizing the control inputs (heating or cooling) in order to minimize a cost function over a receding horizon.
The project is part of a series of experiments with different predictive controllers in the apartment. The access to measurements and actuators is therefore already well established and the student can start their work with relatively little overhead.
If you are a motivated student and are interested in this project, please send an email with a short motivation letter, your transcript of records and CV. Experience with MPC, system identification, building thermal modelling, Python or Matlab are highly beneficial. You will be located at the Urban Energy Systems Laboratory at Empa, your supervising professor will be Prof. John Lygeros of the Automatic Control Laboratory at D-ITET ETH Zürich.
The goal of this semester project is to demonstrate MPC in a real occupied apartment, which is the UMAR Unit in NEST at Empa. This includes the identification of a linear state space representation of the apartment, which can either be done
• by first principles and analysis of architecture plans and used materials,
• with experimental or data-driven system identification techniques, or
• with a combination of these methods.
After identifying the model and setting up the optimization problem, the resulting controller should be implemented on the real system and experiments regarding performance indicators such as user comfort and energy consumption should be conducted.
The goal of this semester project is to demonstrate MPC in a real occupied apartment, which is the UMAR Unit in NEST at Empa. This includes the identification of a linear state space representation of the apartment, which can either be done
• by first principles and analysis of architecture plans and used materials,
• with experimental or data-driven system identification techniques, or
• with a combination of these methods.
After identifying the model and setting up the optimization problem, the resulting controller should be implemented on the real system and experiments regarding performance indicators such as user comfort and energy consumption should be conducted.
Felix Bünning, PhD/Doctoral student, Urban Energy Systems, Empa / Automatic Control Lab, D-ITET, ETH Zurich, felix.buenning@empa.ch
Felix Bünning, PhD/Doctoral student, Urban Energy Systems, Empa / Automatic Control Lab, D-ITET, ETH Zurich, felix.buenning@empa.ch