Ph. D. Project
Title:
Heat pump performances improvement: optimal control under state constraints
Dates:
2025/10/01 - 2028/09/30
Supervisor(s):
Description:
Context
In recent years, heat pumps have become increasingly common in private homes due to their lower energy
costs. They are primarily used for space heating and domestic hot water production. The performance of a
heat pump is mainly defined by two key factors: energy consumption and the level of comfort it provides to
users. To address the challenges of the energy transition, BDR Thermea, a leading manufacturer of heat
pumps, aims to enhance the control systems of its products. The goal is to find the optimal balance between
energy efficiency and user comfort. Recent advances in control theory and artificial intelligence (AI) have
opened up promising new perspectives in this field, which BDR Thermea intends to leverage. To pursue this
goal, BDR Thermea has established a collaboration with the CRAN, LORIA, and LRGP laboratories. This
partnership has been selected for funding by the French Agency for Ecological Transition (ADEME) as part of
the France 2030 initiative.
The main challenge of this project is to develop control strategies that can adapt to different user profiles and
external environmental conditions to minimize energy consumption while maintaining a high level of user
comfort.
The primary objective of the project is to develop advanced control strategies that minimize the energy
consumption of heat pumps while maintaining a high level of user comfort. The research directions are
twofold: one focuses on mathematical modeling of the system, while the other explores data-driven control
techniques (AI). This PhD thesis will primarily investigate the mathematical modeling approach.
Description
In this project, a mathematical model of the heat pump system is already available. This model involves
switching dynamics, reflecting the complex interactions between system components under varying operating
modes. The control of such systems poses significant challenges, especially when aiming to adapt to diverse
user profiles and fluctuating environmental conditions.
To maintain user comfort, the heat pump is supported by a supplementary heating system. This setup adds
another layer of complexity to the control design, as it control allocation between multiple energy sources. The
integration of recent advances in control theory and estimation techniques opens new possibilities for
improving the overall system efficiency and adaptability in real-world scenarios.
Objectives
The objective of this PhD thesis in control engineering are listed next.
- Design an observer capable of extimating the temperature of different layers of water in the tank as well as
the in and out flow of water using only a limited number of temperature sensors.
- Develop an optimal control law that minimizes energy consumption under state constraints.
- Design an optimal control allocation when a supplementary heating system is available.
- Optimize the performances for different user profiles.
- Desgin robust optimal control that account for external disturbances (temperature of the external
environment).
In recent years, heat pumps have become increasingly common in private homes due to their lower energy
costs. They are primarily used for space heating and domestic hot water production. The performance of a
heat pump is mainly defined by two key factors: energy consumption and the level of comfort it provides to
users. To address the challenges of the energy transition, BDR Thermea, a leading manufacturer of heat
pumps, aims to enhance the control systems of its products. The goal is to find the optimal balance between
energy efficiency and user comfort. Recent advances in control theory and artificial intelligence (AI) have
opened up promising new perspectives in this field, which BDR Thermea intends to leverage. To pursue this
goal, BDR Thermea has established a collaboration with the CRAN, LORIA, and LRGP laboratories. This
partnership has been selected for funding by the French Agency for Ecological Transition (ADEME) as part of
the France 2030 initiative.
The main challenge of this project is to develop control strategies that can adapt to different user profiles and
external environmental conditions to minimize energy consumption while maintaining a high level of user
comfort.
The primary objective of the project is to develop advanced control strategies that minimize the energy
consumption of heat pumps while maintaining a high level of user comfort. The research directions are
twofold: one focuses on mathematical modeling of the system, while the other explores data-driven control
techniques (AI). This PhD thesis will primarily investigate the mathematical modeling approach.
Description
In this project, a mathematical model of the heat pump system is already available. This model involves
switching dynamics, reflecting the complex interactions between system components under varying operating
modes. The control of such systems poses significant challenges, especially when aiming to adapt to diverse
user profiles and fluctuating environmental conditions.
To maintain user comfort, the heat pump is supported by a supplementary heating system. This setup adds
another layer of complexity to the control design, as it control allocation between multiple energy sources. The
integration of recent advances in control theory and estimation techniques opens new possibilities for
improving the overall system efficiency and adaptability in real-world scenarios.
Objectives
The objective of this PhD thesis in control engineering are listed next.
- Design an observer capable of extimating the temperature of different layers of water in the tank as well as
the in and out flow of water using only a limited number of temperature sensors.
- Develop an optimal control law that minimizes energy consumption under state constraints.
- Design an optimal control allocation when a supplementary heating system is available.
- Optimize the performances for different user profiles.
- Desgin robust optimal control that account for external disturbances (temperature of the external
environment).
Keywords:
Switching systems, Optimal control, Control allocation, Estimation
Conditions:
Duration: 3 years, starting preferably in the fall of 2025.
Employer: Université de Lorraine (France)
Location: The PhD will take place at Université de Lorraine, CRAN, UMR CNRS 7039 : 2 avenue de
la Forêt de Haye, 54516 Vandoeuvre-lès-Nancy, France.
Funding: 2300 €/month (gross income)
Job profile: We are looking for a candidate holding either a Master degree, an engineering school degree or
any equivalent degree in control engineering or applied mathematics. The candidate should have a strong
background in control theory, mathematical modeling, and numerical simulations. He or she should be able to
work independently and as part of a team, and should have good communication skills in English (French is a
plus). The candidate should be scientifically curious and motivated to do research.
Please contact us by email if you are interested in this PhD position and you need more information.
Employer: Université de Lorraine (France)
Location: The PhD will take place at Université de Lorraine, CRAN, UMR CNRS 7039 : 2 avenue de
la Forêt de Haye, 54516 Vandoeuvre-lès-Nancy, France.
Funding: 2300 €/month (gross income)
Job profile: We are looking for a candidate holding either a Master degree, an engineering school degree or
any equivalent degree in control engineering or applied mathematics. The candidate should have a strong
background in control theory, mathematical modeling, and numerical simulations. He or she should be able to
work independently and as part of a team, and should have good communication skills in English (French is a
plus). The candidate should be scientifically curious and motivated to do research.
Please contact us by email if you are interested in this PhD position and you need more information.
Department(s):
| Control Identification Diagnosis |
Funds:
This doctoral position is supported through funding from the Smartfluid project, financed by ADEME.