Our research focuses on the model & control design, analysis, and optimization of dynamical systems from different domains including robotics and human-machine interaction. It is also important for us to bring control and AI related research into practice by closely cooperating with industry, for instance from the automotive domain, robotics and process automation.
Research projects
Cooperative manipulation with dual-arm robots at the payload limit (headed bei Dr. Andreas Völz)
Kinesthetic teaching and predictive control of interaction tasks in robotics
Distributed model predictive control of nonlinear systems with asynchronous communication
Current projects
Development of an innovative camera-based framework for collision-free human-machine movement
(Third Party Funds Single)
Term: 1. February 2024 - 31. January 2026 Funding source: Bundesministerium für Wirtschaft und Klimaschutz (BMWK)
Hardware architecture, automatic control, autonomy functionality, and developer community: Modular learning control and planning for mobile professional operation vehicles
(Third Party Funds Group – Sub project)
Overall project: POV.OS - Hardware and software platform for mobile professional operation vehicles Term: 1. January 2023 - 31. December 2025 Funding source: Bundesministerium für Wirtschaft und Klimaschutz (BMWK)
Formulation of dispersed systems via (melt) emulsification: Process design, in situ diagnostics and regulation
(Third Party Funds Group – Sub project)
Overall project: Autonome Prozesse in der Partikeltechnik - Erforschung und Erprobung von Konzepten zur modellbasierten Führung partikeltechnischer Prozesse Term: 1. January 2023 - 31. December 2025 Funding source: DFG / Schwerpunktprogramm (SPP)
The aim of this project is the automated production of liquid-liquid disperse systems via melt emulsification, whereby in this process emulsification takes place at elevated temperature. The products obtained after cooling are dispersions of spherical nanoparticles or microparticles. Within the scope of this project, a melt emulsification device for the automated production of product particles with a well-defined particle size distribution (PSD) will be further developed. The PSD has a significant influence on the subsequent product properties, such as flow behavior or drug release kinetics. The PSD of the products is determined by the complex interaction of competing mechanisms. These are, in particular, droplet breakup in a rotor-stator device as a result of shear and elongation stress, as well as coalescence and further ripening, which in turn depend on the system composition, i.e. the emulsifier used (type, concentration) and the dispersion phase (viscosity, volume fraction).
Therefore, for a better process understanding and an active process control, possibilities for in situ determination of the PSD are urgently required. In this project, a novel fiber-coupled measurement system based on broadband elastic light scattering is developed for in situ measurement of the PSD. The system will be validated on reference particle systems and applied to the emulsification process. Furthermore, a hybrid process model is developed, which is the basis for the design of a model predictive control of the process. The model predictive control in combination with the in situ measurement will provide the possibility for an active process control and the production of emulsions with predefined properties and a simultaneous optimization of the process time.
Formulierung von dispersen Systemen durch (Schmelz-)Emulgierung: Prozessgestaltung, In-situ-Diagnostik und Regelung
(Third Party Funds Group – Sub project)
Overall project: Autonome Prozesse in der Partikeltechnik - Erforschung und Erprobung von Konzepten zur modellbasierten Führung partikeltechnischer Prozesse Term: 1. January 2023 - 31. December 2025 Funding source: DFG / Schwerpunktprogramm (SPP)
Ziel dieses Projekts ist die automatisierte Herstellung von flüssig-flüssig-dispersen Systemen über Schmelzeemulgieren, wobei bei diesem Prozess das Emulgieren bei erhöhter Temperatur erfolgt. Als Produkte werden nach der Abkühlung Dispersionen von sphärischen Nano- oder Mikropartikeln erhalten. In Rahmen dieses Projekts wird ein Schmelzeemulgierprozess für die automatisierte Herstellung von Produktpartikeln mit wohldefinierter Partikelgrößenverteilung (PGV) betrachtet. Diese beeinflusst dabei maßgeblich die späteren Produkteigenschaften, wie zum Beispiel das Fließverhalten oder wie die Wirkstofffreisetzungskinetik. Die PGV der Produkte wird dabei durch das komplexe Zusammenspiel konkurrierender Mechanismen bestimmt. Dies sind insbesondere der Tropfenaufbruch in einem Rotor-Stator infolge von Scher- und Dehnbeanspruchung sowie die Koaleszenz und weitere Reifung, die ihrerseits von der Systemzusammensetzung, d.h. dem genutzten Emulgator (Art, Konzentration) und der Dispersphase (Viskosität, Volumenanteil) abhängig sind. Für ein besseres Prozessverständnis und eine aktive Prozessregelung sind daher Möglichkeiten zur in situ Bestimmung der PGV dringend erforderlich. In diesem Projekt wird zur in situ Messung der PGV ein neuartiges, auf breitbandiger elastischer Lichtstreuung basierendes fasergekoppeltes Messsystem entwickelt. Dieses wird an Referenzpartikelsystemen validiert und am Emulgierprozess eingesetzt. Weiterhin wird ein hybrides Prozessmodell entwickelt, das die Basis für das Design einer modellprädiktiven Regelung des Prozesses darstellt. Die modellprädiktive Regelung wird in Kombination mit der in situ Messung die Möglichkeit für eine aktive Prozesssteuerung und die Herstellung von Emulsionen mit vorher definierten Eigenschaften bei gleichzeitiger Optimierung der Prozesszeit ermöglichen.
Overall project: AGENT-2: Agent-based data-driven modeling for stochastic and self-adjusting control of building energy systems Term: 1. November 2022 - 31. October 2025 Funding source: Bundesministerium für Wirtschaft und Klimaschutz (BMWK)
To achieve climate targets, CO2 emissions in the building sector have to be significantly reduced. However, the integration of renewable energy sources increases the complexity of building energy systems and thus the requirements for the operation strategy. Model-based and predictive controllers are necessary for efficient operation. However, due to the high complexity of the energy systems, the development, implementation, and commissioning are very complex leading to high costs, which is why model predictive and optimization-based control strategies are rarely used in practice so far. The goal of the AGENT-2 project is to develop a self-adjusting and self-learning model-predictive control concept that reduces the implementation and commissioning effort and thus increases the applicability of efficient operating strategies in practice. The control concept to be developed is based on distributed agents, each of which learns the system behavior of a subsystem and controls the subsystem. This is based on the findings and the framework developed in the previous project AGENT. The operation of the overall system is achieved by the interaction oft h e self-learning agents with each other. Thus, a self-adjusting and scalable control strategy for building energy systems is created. The self-learning control strategy is compared with state-of-the-art concepts in simulations and tested in practical operation in two demonstration buildings. The findings will be generalized and possibilities for the transfer into practice will be investigated. The project thus contributes to increasing the efficiency of building operation and to reducing the costs of controller implementation and commissioning.
Robust Planning and Control using Probabilistic Methods
(Third Party Funds Group – Sub project)
Overall project: Verbundprojekt MANNHEIM-AUTOtech.agil: Architektur und Technologien zur Orchestrierung automobiltechnischer Agilität Term: 1. October 2022 - 30. September 2025 Funding source: Bundesministerium für Bildung und Forschung (BMBF)
Precise interactions as part of industrial manufacturing tasks are typically very complex to characterize and implement. One reason for this is the heterogeneity of the task-specific requirements for the motion and control behavior. A direct implementation of the task into a robot program therefore requires highly qualified specialists and is only profitable for large lot sizes. For a flexible applicability and easy (re-)configuration of the robot system, an approach to programming by kinesthetic demonstration is developed in this project. The robot is guided by the user through the entire manipulation task, while the robot motion as well as the interaction forces are simultaneously recorded. Typically, several repetitions of the demonstration are necessary in order to compensate for the suboptimality and imprecision of the human demonstration. This is particularly important for complex motion sequences or interaction situations, such as periodic movements or the assembly of components, that are difficult to demonstrate but at the same time are crucial for a successful task execution.
The basis for this project is a previously developed general framework for model predictive interaction control (MPIC). The manipulation task is split into a sequence of elementary tasks, so-called manipulation primitives (MPs) with individual motion and control characteristics, which are treated in a holistic manner by a model predictive control approach. The MPIC approach is elaborated in this project regarding the kinesthetic demonstration of manipulation tasks, e.g. by considering the switching between MPs over the prediction horizon of the MPC. A further focus lies on the automatic generation of the MP sequence from the repeated demonstration of the manipulation task without requiring additional expert knowledge. Based on the demonstration, the manipulation task will be iteratively refined by learning the setpoints and the transition conditions of the MPs and finally by optimizing the overall manipulation task.
Innovative Regelungs- und Steuerungsstrategien für Druckerhöhungsanlagen - TP Erlangen
(Third Party Funds Group – Sub project)
Overall project: Innovative Regelungs- und Steuerungsstrategien für Druckerhöhungsanlagen Term: 1. July 2021 - 30. June 2024 Funding source: Bayerische Forschungsstiftung
Kißkalt, J., Michalka, A., Strohmeyer, C., Horn, M., & Graichen, K. (2023). Simulation chain for sensorized strain wave gears. In Proc. 27th International Conference on System Theory, Control and Computing (ICSTCC) (pp. 467 - 473). Timisoara (Romania).
Snobar, F., Reinhard, J., Huber, H., Hoffmann, M., Stelzig, M., Vossiek, M., & Graichen, K. (2022). FOV-based model predictive object tracking for quadcopters. In Proceedings of the 9th IFAC Symposium on Mechatronic Systems (Mechatronics 2022) (pp. 13 - 18). Los Angeles, CA (USA).
Bergmann, D., & Graichen, K. (2020). Safe Bayesian Optimization under Unknown Constraints. In 59th IEEE Conference on Decision and Control (CDC 2020) (pp. 3592-3597). Institute of Electrical and Electronics Engineers Inc..
Our research focuses on the model & control design, analysis, and optimization of dynamical systems from different domains including robotics and human-machine interaction. It is also important for us to bring control and AI related research into practice by closely cooperating with industry, for instance from the automotive domain, robotics and process automation.
Research projects
Current projects
Development of an innovative camera-based framework for collision-free human-machine movement
(Third Party Funds Single)
Funding source: Bundesministerium für Wirtschaft und Klimaschutz (BMWK)
Robust Reinforcement Learning for Thermal Management Control
(Third Party Funds Single)
Funding source: Industrie
Model predictive flight control
(Third Party Funds Single)
Funding source: Industrie
Robust energy-based control of MMC/HVDV systems
(Third Party Funds Single)
Funding source: Industrie
Hardware architecture, automatic control, autonomy functionality, and developer community: Modular learning control and planning for mobile professional operation vehicles
(Third Party Funds Group – Sub project)
Term: 1. January 2023 - 31. December 2025
Funding source: Bundesministerium für Wirtschaft und Klimaschutz (BMWK)
Formulation of dispersed systems via (melt) emulsification: Process design, in situ diagnostics and regulation
(Third Party Funds Group – Sub project)
Term: 1. January 2023 - 31. December 2025
Funding source: DFG / Schwerpunktprogramm (SPP)
The aim of this project is the automated production of liquid-liquid disperse systems via melt emulsification, whereby in this process emulsification takes place at elevated temperature. The products obtained after cooling are dispersions of spherical nanoparticles or microparticles. Within the scope of this project, a melt emulsification device for the automated production of product particles with a well-defined particle size distribution (PSD) will be further developed. The PSD has a significant influence on the subsequent product properties, such as flow behavior or drug release kinetics. The PSD of the products is determined by the complex interaction of competing mechanisms. These are, in particular, droplet breakup in a rotor-stator device as a result of shear and elongation stress, as well as coalescence and further ripening, which in turn depend on the system composition, i.e. the emulsifier used (type, concentration) and the dispersion phase (viscosity, volume fraction).
Therefore, for a better process understanding and an active process control, possibilities for in situ determination of the PSD are urgently required. In this project, a novel fiber-coupled measurement system based on broadband elastic light scattering is developed for in situ measurement of the PSD. The system will be validated on reference particle systems and applied to the emulsification process. Furthermore, a hybrid process model is developed, which is the basis for the design of a model predictive control of the process. The model predictive control in combination with the in situ measurement will provide the possibility for an active process control and the production of emulsions with predefined properties and a simultaneous optimization of the process time.
Formulierung von dispersen Systemen durch (Schmelz-)Emulgierung: Prozessgestaltung, In-situ-Diagnostik und Regelung
(Third Party Funds Group – Sub project)
Term: 1. January 2023 - 31. December 2025
Funding source: DFG / Schwerpunktprogramm (SPP)
Ziel dieses Projekts ist die automatisierte Herstellung von flüssig-flüssig-dispersen Systemen über Schmelzeemulgieren, wobei bei diesem Prozess das Emulgieren bei erhöhter Temperatur erfolgt. Als Produkte werden nach der Abkühlung Dispersionen von sphärischen Nano- oder Mikropartikeln erhalten. In Rahmen dieses Projekts wird ein Schmelzeemulgierprozess für die automatisierte Herstellung von Produktpartikeln mit wohldefinierter Partikelgrößenverteilung (PGV) betrachtet. Diese beeinflusst dabei maßgeblich die späteren Produkteigenschaften, wie zum Beispiel das Fließverhalten oder wie die Wirkstofffreisetzungskinetik. Die PGV der Produkte wird dabei durch das komplexe Zusammenspiel konkurrierender Mechanismen bestimmt. Dies sind insbesondere der Tropfenaufbruch in einem Rotor-Stator infolge von Scher- und Dehnbeanspruchung sowie die Koaleszenz und weitere Reifung, die ihrerseits von der Systemzusammensetzung, d.h. dem genutzten Emulgator (Art, Konzentration) und der Dispersphase (Viskosität, Volumenanteil) abhängig sind. Für ein besseres Prozessverständnis und eine aktive Prozessregelung sind daher Möglichkeiten zur in situ Bestimmung der PGV dringend erforderlich. In diesem Projekt wird zur in situ Messung der PGV ein neuartiges, auf breitbandiger elastischer Lichtstreuung basierendes fasergekoppeltes Messsystem entwickelt. Dieses wird an Referenzpartikelsystemen validiert und am Emulgierprozess eingesetzt. Weiterhin wird ein hybrides Prozessmodell entwickelt, das die Basis für das Design einer modellprädiktiven Regelung des Prozesses darstellt. Die modellprädiktive Regelung wird in Kombination mit der in situ Messung die Möglichkeit für eine aktive Prozesssteuerung und die Herstellung von Emulsionen mit vorher definierten Eigenschaften bei gleichzeitiger Optimierung der Prozesszeit ermöglichen.
Predictive and learning control methods
(Third Party Funds Group – Sub project)
Term: 1. November 2022 - 31. October 2025
Funding source: Bundesministerium für Wirtschaft und Klimaschutz (BMWK)
To achieve climate targets, CO2 emissions in the building sector have to be significantly reduced. However, the integration of renewable energy sources increases the complexity of building energy systems and thus the requirements for the operation strategy. Model-based and predictive controllers are necessary for efficient operation. However, due to the high complexity of the energy systems, the development, implementation, and commissioning are very complex leading to high costs, which is why model predictive and optimization-based control strategies are rarely used in practice so far. The goal of the AGENT-2 project is to develop a self-adjusting and self-learning model-predictive control concept that reduces the implementation and commissioning effort and thus increases the applicability of efficient operating strategies in practice. The control concept to be developed is based on distributed agents, each of which learns the system behavior of a subsystem and controls the subsystem. This is based on the findings and the framework developed in the previous project AGENT. The operation of the overall system is achieved by the interaction oft h e self-learning agents with each other. Thus, a self-adjusting and scalable control strategy for building energy systems is created. The self-learning control strategy is compared with state-of-the-art concepts in simulations and tested in practical operation in two demonstration buildings. The findings will be generalized and possibilities for the transfer into practice will be investigated. The project thus contributes to increasing the efficiency of building operation and to reducing the costs of controller implementation and commissioning.
Robust Planning and Control using Probabilistic Methods
(Third Party Funds Group – Sub project)
Term: 1. October 2022 - 30. September 2025
Funding source: Bundesministerium für Bildung und Forschung (BMBF)
Control of ring resonator modulators in optical communication
(Third Party Funds Single)
Funding source: Industrie
Distributed model predictive control of nonlinear systems with asynchronous communication
(Third Party Funds Single)
Funding source: Deutsche Forschungsgemeinschaft (DFG)
Anomaly detection and intelligent recalibration of sensorized systems
(Third Party Funds Single)
Funding source: Industrie
KI-unterstützte Modellierung zur Steigerung der Regelgüte
(Third Party Funds Single)
Funding source: Industrie
Kinesthetic teaching and predictive control of interaction tasks in robotics
(Third Party Funds Single)
Funding source: Deutsche Forschungsgemeinschaft (DFG)
Precise interactions as part of industrial manufacturing tasks are typically very complex to characterize and implement. One reason for this is the heterogeneity of the task-specific requirements for the motion and control behavior. A direct implementation of the task into a robot program therefore requires highly qualified specialists and is only profitable for large lot sizes. For a flexible applicability and easy (re-)configuration of the robot system, an approach to programming by kinesthetic demonstration is developed in this project. The robot is guided by the user through the entire manipulation task, while the robot motion as well as the interaction forces are simultaneously recorded. Typically, several repetitions of the demonstration are necessary in order to compensate for the suboptimality and imprecision of the human demonstration. This is particularly important for complex motion sequences or interaction situations, such as periodic movements or the assembly of components, that are difficult to demonstrate but at the same time are crucial for a successful task execution.
The basis for this project is a previously developed general framework for model predictive interaction control (MPIC). The manipulation task is split into a sequence of elementary tasks, so-called manipulation primitives (MPs) with individual motion and control characteristics, which are treated in a holistic manner by a model predictive control approach. The MPIC approach is elaborated in this project regarding the kinesthetic demonstration of manipulation tasks, e.g. by considering the switching between MPs over the prediction horizon of the MPC. A further focus lies on the automatic generation of the MP sequence from the repeated demonstration of the manipulation task without requiring additional expert knowledge. Based on the demonstration, the manipulation task will be iteratively refined by learning the setpoints and the transition conditions of the MPs and finally by optimizing the overall manipulation task.
Innovative Regelungs- und Steuerungsstrategien für Druckerhöhungsanlagen - TP Erlangen
(Third Party Funds Group – Sub project)
Term: 1. July 2021 - 30. June 2024
Funding source: Bayerische Forschungsstiftung
Robust control of modular multi-level converters
(Third Party Funds Single)
Funding source: Industrie
Recent publications
2024
2023
2022
2021
2020
Related Research Fields
Contact: