Register now After registration you will be able to apply for this opportunity online.
This opportunity is not published. No applications will be accepted.
Tuning of data-driven controllers for building climate control
Energy consumption in buildings accounts for over 30% of worldwide energy consumption. Mitigating energy consumption is typically done by improving operation of heating and cooling systems in buildings. However, there is limited analysis of implementation and tuning of data-driven controllers, which prevents their wide use. This project addresses the challenges related to tuning of data-driven controllers for implementation.
Keywords: Data-enabled predictive control, sensitivity analysis, data quality
Energy consumption in buildings accounts for over 30% of worldwide energy consumption. Mitigating energy consumption is typically done by improving operation of heating and cooling systems in buildings. Efficient operation of heating and cooling is a part of climate control in buildings. The main task of climate control is to satisfy heating and cooling needs of the occupants of buildings using minimal amount of energy. Due to the inherent complexity of the requirements and the dynamics of buildings, developing efficient climate controllers is a challenge. Data-driven controllers, such as DeePC, have a great potential thanks to their simplicity and ability to work with limited knowledge of the controlled building. However, there is limited analysis of implementation and tuning of these data-driven controllers, which prevents their wide use. This project addresses the challenges related to tuning of data-driven controllers for implementation.
Implementation of any controller in practical applications requires tuning of parameters of the controller. For instance, the parameters Kp and Ki must be chosen if a PI controller is to be used. The choice of these parameters can be done by doing tests on the system or by solving an optimisation problem. Model-based controllers, such as MPC, can also be tuned because the influence of the parameters of the controller on the performance of the system is known. However, novel data-based and model-free controllers have not yet been analysed from the perspective of the effect of their tuning on the performance.
Energy consumption in buildings accounts for over 30% of worldwide energy consumption. Mitigating energy consumption is typically done by improving operation of heating and cooling systems in buildings. Efficient operation of heating and cooling is a part of climate control in buildings. The main task of climate control is to satisfy heating and cooling needs of the occupants of buildings using minimal amount of energy. Due to the inherent complexity of the requirements and the dynamics of buildings, developing efficient climate controllers is a challenge. Data-driven controllers, such as DeePC, have a great potential thanks to their simplicity and ability to work with limited knowledge of the controlled building. However, there is limited analysis of implementation and tuning of these data-driven controllers, which prevents their wide use. This project addresses the challenges related to tuning of data-driven controllers for implementation.
Implementation of any controller in practical applications requires tuning of parameters of the controller. For instance, the parameters Kp and Ki must be chosen if a PI controller is to be used. The choice of these parameters can be done by doing tests on the system or by solving an optimisation problem. Model-based controllers, such as MPC, can also be tuned because the influence of the parameters of the controller on the performance of the system is known. However, novel data-based and model-free controllers have not yet been analysed from the perspective of the effect of their tuning on the performance.
The project focuses on DeePC and applies it to building climate control. The controller has been shown to perform well in practice. This project aims to answer the following questions:
- How do the parameters of DeePC affect the performance of the system?
- How do the properties of data affect the performance of the system and the controller?
- What theoretical properties does DeePC have with respect to its hyperparameters?
The expect result of the project is a set of guidelines for tuning hyperparameters of DeePC for building climate control. If successful, the project will finish with validation of the proposed tuning guidelines in NEST building at EMPA.
The project focuses on DeePC and applies it to building climate control. The controller has been shown to perform well in practice. This project aims to answer the following questions: - How do the parameters of DeePC affect the performance of the system? - How do the properties of data affect the performance of the system and the controller? - What theoretical properties does DeePC have with respect to its hyperparameters?
The expect result of the project is a set of guidelines for tuning hyperparameters of DeePC for building climate control. If successful, the project will finish with validation of the proposed tuning guidelines in NEST building at EMPA.
Please send your CV and transcript of records in PDF format via email to mbadyn@control.ee.ethz.ch and mzagorowska@control.ee.ethz.ch
Please send your CV and transcript of records in PDF format via email to mbadyn@control.ee.ethz.ch and mzagorowska@control.ee.ethz.ch