-
SafEDMD: A certified learning architecture tailored to data-driven control of nonlinear dynamical systems
Authors:
Robin Strässer,
Manuel Schaller,
Karl Worthmann,
Julian Berberich,
Frank Allgöwer
Abstract:
The Koopman operator serves as the theoretical backbone for machine learning of dynamical control systems, where the operator is heuristically approximated by extended dynamic mode decomposition (EDMD). In this paper, we propose Stability- and certificate-oriented EDMD (SafEDMD): a novel EDMD-based learning architecture which comes along with rigorous certificates, resulting in a reliable surrogat…
▽ More
The Koopman operator serves as the theoretical backbone for machine learning of dynamical control systems, where the operator is heuristically approximated by extended dynamic mode decomposition (EDMD). In this paper, we propose Stability- and certificate-oriented EDMD (SafEDMD): a novel EDMD-based learning architecture which comes along with rigorous certificates, resulting in a reliable surrogate model generated in a data-driven fashion. To ensure the trustworthiness of SafEDMD, we derive proportional error bounds, which vanish at the origin and are tailored to control tasks, leading to certified controller design based on semi-definite programming. We illustrate the developed method by means of several benchmark examples and highlight the advantages over state-of-the-art methods.
△ Less
Submitted 17 May, 2024; v1 submitted 5 February, 2024;
originally announced February 2024.
-
Jet space extensions of infinite-dimensional Hamiltonian systems
Authors:
Till Preuster,
Manuel Schaller,
Bernhard Maschke
Abstract:
We analyze infinite-dimensional Hamiltonian systems corresponding to partial differential equations on one-dimensional spatial domains formulated with formally skew-adjoint Hamiltonian operators and nonlinear Hamiltonian density. In various applications, the Hamiltonian density can depend on spatial derivatives of the state such that these systems can not straightforwardly be formulated as boundar…
▽ More
We analyze infinite-dimensional Hamiltonian systems corresponding to partial differential equations on one-dimensional spatial domains formulated with formally skew-adjoint Hamiltonian operators and nonlinear Hamiltonian density. In various applications, the Hamiltonian density can depend on spatial derivatives of the state such that these systems can not straightforwardly be formulated as boundary port-Hamiltonian system using a Stokes-Dirac structure. In this work, we show that any Hamiltonian system of the above class can be reformulated as a Hamiltonian system on the jet space, in which the Hamiltonian density only depends on the extended state variable itself and not on its derivatives. Consequently, well-known geometric formulations with Stokes- Dirac structures are applicable. Additionally, we provide a similar result for dissipative systems. We illustrate the developed theory by means of the the Boussinesq equation, the dynamics of an elastic rod and the Allen-Cahn equation.
△ Less
Submitted 24 January, 2024;
originally announced January 2024.
-
Koopman-based feedback design with stability guarantees
Authors:
Robin Strässer,
Manuel Schaller,
Karl Worthmann,
Julian Berberich,
Frank Allgöwer
Abstract:
We present a method to design a state-feedback controller ensuring exponential stability for nonlinear systems using only measurement data. Our approach relies on Koopman-operator theory and uses robust control to explicitly account for approximation errors due to finitely many data samples. To simplify practical usage across various applications, we provide a tutorial-style exposition of the feed…
▽ More
We present a method to design a state-feedback controller ensuring exponential stability for nonlinear systems using only measurement data. Our approach relies on Koopman-operator theory and uses robust control to explicitly account for approximation errors due to finitely many data samples. To simplify practical usage across various applications, we provide a tutorial-style exposition of the feedback design and its stability guarantees for single-input systems. Moreover, we extend this controller design to multi-input systems and more flexible nonlinear state-feedback controllers using gain-scheduling techniques to increase the guaranteed region of attraction. As the proposed controller design is framed as a semidefinite program, it allows for an efficient solution. Further, we enhance the geometry of the region of attraction through a heuristic algorithm that establishes a connection between the employed Koopman lifting and the dynamics of the system. Finally, we validate the proposed feedback design procedure by means of numerical examples.
△ Less
Submitted 3 May, 2024; v1 submitted 3 December, 2023;
originally announced December 2023.
-
Data-driven MPC with stability guarantees using extended dynamic mode decomposition
Authors:
Lea Bold,
Lars Grüne,
Manuel Schaller,
Karl Worthmann
Abstract:
For nonlinear (control) systems, extended dynamic mode decomposition (EDMD) is a popular method to obtain data-driven surrogate models. Its theoretical foundation is the Koopman framework, in which one propagates observable functions of the state to obtain a linear representation in an infinite-dimensional space. In this work, we prove practical asymptotic stability of a (controlled) equilibrium f…
▽ More
For nonlinear (control) systems, extended dynamic mode decomposition (EDMD) is a popular method to obtain data-driven surrogate models. Its theoretical foundation is the Koopman framework, in which one propagates observable functions of the state to obtain a linear representation in an infinite-dimensional space. In this work, we prove practical asymptotic stability of a (controlled) equilibrium for EDMD-based model predictive control, in which the optimization step is conducted using the data-based surrogate model. To this end, we derive novel bounds on the estimation error that are proportional to the norm of state and control. This enables us to show that, if the underlying system is cost controllable, this stabilizablility property is preserved. We conduct numerical simulations illustrating the proven practical asymptotic stability.
△ Less
Submitted 11 March, 2024; v1 submitted 1 August, 2023;
originally announced August 2023.
-
Reprojection methods for Koopman-based modelling and prediction
Authors:
Pieter van Goor,
Robert Mahony,
Manuel Schaller,
Karl Worthmann
Abstract:
Extended Dynamic Mode Decomposition (eDMD) is a powerful tool to generate data-driven surrogate models for the prediction and control of nonlinear dynamical systems in the Koopman framework. In eDMD a compression of the lifted system dynamics on the space spanned by finitely many observables is computed, in which the original space is embedded as a low-dimensional manifold. While this manifold is…
▽ More
Extended Dynamic Mode Decomposition (eDMD) is a powerful tool to generate data-driven surrogate models for the prediction and control of nonlinear dynamical systems in the Koopman framework. In eDMD a compression of the lifted system dynamics on the space spanned by finitely many observables is computed, in which the original space is embedded as a low-dimensional manifold. While this manifold is invariant for the infinite-dimensional Koopman operator, this invariance is typically not preserved for its eDMD-based approximation. Hence, an additional (re-)projection step is often tacitly incorporated to improve the prediction capability. We propose a novel framework for consistent reprojectors respecting the underlying manifold structure. Further, we present a new geometric reprojector based on maximum-likelihood arguments, which significantly enhances the approximation accuracy and preserves known finite-data error bounds.
△ Less
Submitted 30 July, 2023;
originally announced July 2023.
-
Energy-optimal control of adaptive structures
Authors:
Manuel Schaller,
Amelie Zeller,
Michael Böhm,
Oliver Sawodny,
Cristina Tarín,
Karl Worthmann
Abstract:
Adaptive structures are equipped with sensors and actuators to actively counteract external loads such as wind. This can significantly reduce resource consumption and emissions during the life cycle compared to conventional structures. A common approach for active dam** is to derive a port-Hamiltonian model and to employ linear-quadratic control. However, the quadratic control penalization lacks…
▽ More
Adaptive structures are equipped with sensors and actuators to actively counteract external loads such as wind. This can significantly reduce resource consumption and emissions during the life cycle compared to conventional structures. A common approach for active dam** is to derive a port-Hamiltonian model and to employ linear-quadratic control. However, the quadratic control penalization lacks physical interpretation and merely serves as a regularization term. Rather, we propose a controller, which achieves the goal of vibration dam** while acting energy-optimal. Leveraging the port-Hamiltonian structure, we show that the optimal control is uniquely determined, even on singular arcs. Further, we prove a stable long-time behavior of optimal trajectories by means of a turnpike property. Last, the proposed controller's efficiency is evaluated in a numerical study.
△ Less
Submitted 8 December, 2023; v1 submitted 23 June, 2023;
originally announced June 2023.
-
Optimal control of port-Hamiltonian systems: energy, entropy, and exergy
Authors:
Friedrich Philipp,
Manuel Schaller,
Karl Worthmann,
Timm Faulwasser,
Bernhard Maschke
Abstract:
We consider irreversible and coupled reversible-irreversible nonlinear port-Hamiltonian systems and the respective sets of thermodynamic equilibria. In particular, we are concerned with optimal state transitions and output stabilization on finite-time horizons. We analyze a class of optimal control problems, where the performance functional can be interpreted as a linear combination of energy supp…
▽ More
We consider irreversible and coupled reversible-irreversible nonlinear port-Hamiltonian systems and the respective sets of thermodynamic equilibria. In particular, we are concerned with optimal state transitions and output stabilization on finite-time horizons. We analyze a class of optimal control problems, where the performance functional can be interpreted as a linear combination of energy supply, entropy generation, or exergy supply. Our results establish the integral turnpike property towards the set of thermodynamic equilibria providing a rigorous connection of optimal system trajectories to optimal steady states. Throughout the paper, we illustrate our findings by means of two examples: a network of heat exchangers and a gas-piston system.
△ Less
Submitted 15 June, 2023;
originally announced June 2023.
-
Structured Optimization-Based Model Order Reduction for Parametric Systems
Authors:
Paul Schwerdtner,
Manuel Schaller
Abstract:
We develop an optimization-based algorithm for parametric model order reduction (PMOR) of linear time-invariant dynamical systems. Our method aims at minimizing the $\mathcal{H}_\infty \otimes \mathcal{L}_\infty$ approximation error in the frequency and parameter domain by an optimization of the reduced order model (ROM) matrices. State-of-the-art PMOR methods often compute several nonparametric R…
▽ More
We develop an optimization-based algorithm for parametric model order reduction (PMOR) of linear time-invariant dynamical systems. Our method aims at minimizing the $\mathcal{H}_\infty \otimes \mathcal{L}_\infty$ approximation error in the frequency and parameter domain by an optimization of the reduced order model (ROM) matrices. State-of-the-art PMOR methods often compute several nonparametric ROMs for different parameter samples, which are then combined to a single parametric ROM. However, these parametric ROMs can have a low accuracy between the utilized sample points. In contrast, our optimization-based PMOR method minimizes the approximation error across the entire parameter domain. Moreover, due to our flexible approach of optimizing the system matrices directly, we can enforce favorable features such as a port-Hamiltonian structure in our ROMs across the entire parameter domain.
Our method is an extension of the recently developed SOBMOR-algorithm to parametric systems. We extend both the ROM parameterization and the adaptive sampling procedure to the parametric case. Several numerical examples demonstrate the effectiveness and high accuracy of our method in a comparison with other PMOR methods.
△ Less
Submitted 20 March, 2023; v1 submitted 12 September, 2022;
originally announced September 2022.
-
State and parameter estimation for retinal laser treatment
Authors:
Viktoria Kleyman,
Manuel Schaller,
Mario Mordmuller,
Mitsuru Wilson,
Ralf Brinkmann,
Karl Worthmann,
Matthias A. Muller
Abstract:
Adequate therapeutic retinal laser irradiation needs to be adapted to the local absorption. This leads to time-consuming treatments as the laser power needs to be successively adjusted to avoid under- and overtreatment caused by too low or too high temperatures. Closed-loop control can overcome this burden by means of temperature measurements. To allow for model predictive control schemes, the cur…
▽ More
Adequate therapeutic retinal laser irradiation needs to be adapted to the local absorption. This leads to time-consuming treatments as the laser power needs to be successively adjusted to avoid under- and overtreatment caused by too low or too high temperatures. Closed-loop control can overcome this burden by means of temperature measurements. To allow for model predictive control schemes, the current state and the spot-dependent absorption need to be estimated. In this paper, we thoroughly compare moving horizon estimator (MHE) and extended Kalman filter (EKF) designs for joint state and parameter estimation. We consider two different scenarios, the estimation of one or two unknown absorption coefficients. For one unknown parameter, both estimators perform very similar. For two unknown parameters, we found that the MHE benefits from active parameter constraints at the beginning of the estimation, whereas after a settling time both estimators perform again very similar as long as the parameters are inside the considered parameter bounds.
△ Less
Submitted 23 March, 2022;
originally announced March 2022.
-
An Adaptive Human Driver Model for Realistic Race Car Simulations
Authors:
Stefan Löckel,
Siwei Ju,
Maximilian Schaller,
Peter van Vliet,
Jan Peters
Abstract:
Engineering a high-performance race car requires a direct consideration of the human driver using real-world tests or Human-Driver-in-the-Loop simulations. Apart from that, offline simulations with human-like race driver models could make this vehicle development process more effective and efficient but are hard to obtain due to various challenges. With this work, we intend to provide a better und…
▽ More
Engineering a high-performance race car requires a direct consideration of the human driver using real-world tests or Human-Driver-in-the-Loop simulations. Apart from that, offline simulations with human-like race driver models could make this vehicle development process more effective and efficient but are hard to obtain due to various challenges. With this work, we intend to provide a better understanding of race driver behavior and introduce an adaptive human race driver model based on imitation learning. Using existing findings and an interview with a professional race engineer, we identify fundamental adaptation mechanisms and how drivers learn to optimize lap time on a new track. Subsequently, we use these insights to develop generalization and adaptation techniques for a recently presented probabilistic driver modeling approach and evaluate it using data from professional race drivers and a state-of-the-art race car simulator. We show that our framework can create realistic driving line distributions on unseen race tracks with almost human-like performance. Moreover, our driver model optimizes its driving lap by lap, correcting driving errors from previous laps while achieving faster lap times. This work contributes to a better understanding and modeling of the human driver, aiming to expedite simulation methods in the modern vehicle development process and potentially supporting automated driving and racing technologies.
△ Less
Submitted 20 July, 2022; v1 submitted 3 March, 2022;
originally announced March 2022.
-
Parameter estimation and model reduction for retinal laser treatment
Authors:
Manuel Schaller,
Mitsuru Wilson,
Viktoria Kleyman,
Mario Mordmüller,
Ralf Brinkmann,
Matthias A. Müller,
Karl Worthmann
Abstract:
Laser photocoagulation is one of the most frequently used treatment approaches for retinal diseases such as diabetic retinopathy and macular edema. The use of model-based control, such as Model Predictive Control (MPC), enhances a safe and effective treatment by guaranteeing temperature bounds. In general, real-time requirements for model-based control designs are not met since the temperature dis…
▽ More
Laser photocoagulation is one of the most frequently used treatment approaches for retinal diseases such as diabetic retinopathy and macular edema. The use of model-based control, such as Model Predictive Control (MPC), enhances a safe and effective treatment by guaranteeing temperature bounds. In general, real-time requirements for model-based control designs are not met since the temperature distribution in the eye fundus is governed by a heat equation with a nonlinear parameter dependency. This issue is circumvented by representing the model by a lower-dimensional system which well-approximates the original model, including the parametric dependency. We combine a global-basis approach with the discrete empirical interpolation method, tailor its hyperparameters to laser photocoagulation, and show its superiority in comparison to a recently proposed method based on Taylor-series approximation. Its effectiveness is measured in computation time for MPC. We further present a case study to estimate the range of absorption parameters in porcine eyes, and by means of a theoretical and numerical sensitivity analysis we show that the sensitivity of the temperature increase is higher with respect to the absorption coefficient of the retinal pigment epithelium (RPE) than of the choroid's.
△ Less
Submitted 25 February, 2022;
originally announced February 2022.
-
Model predictive control for retinal laser treatment at 1 kHz
Authors:
Manuel Schaller,
Viktoria Kleyman,
Mario Mordmüller,
Christian Schmidt,
Mitsuru Wilson,
Ralf Brinkmann,
Matthias A. Müller,
Karl Worthmann
Abstract:
Laser photocoagulation is a technique applied in the treatment of retinal diseases. While this is often done manually or using simple control schemes, we pursue an optimization-based approach, namely Model Predictive Control (MPC), to enforce bounds on the peak temperature and, thus, safety during the medical treatment procedure - despite the spot-dependent absorption of the tissue. To this end, a…
▽ More
Laser photocoagulation is a technique applied in the treatment of retinal diseases. While this is often done manually or using simple control schemes, we pursue an optimization-based approach, namely Model Predictive Control (MPC), to enforce bounds on the peak temperature and, thus, safety during the medical treatment procedure - despite the spot-dependent absorption of the tissue. To this end, a repetition rate of 1 kHz is desirable rendering the real-time requirements a major challenge. We present a tailored MPC scheme using parametric model reduction, an extended Kalman filter for the parameter and state estimation, and suitably constructed stage costs and verify its applicability both in simulation and experiments with porcine eyes. Moreover, we give some insight on the implementation specifically tailored for fast numerical computations.
△ Less
Submitted 24 March, 2022; v1 submitted 25 February, 2022;
originally announced February 2022.
-
State and parameter estimation for model-based retinal laser treatment
Authors:
Viktoria Kleyman,
Manuel Schaller,
Mitsuru Wilson,
Mario Mordmüller,
Ralf Brinkmann,
Karl Worthmann,
Matthias A. Müller
Abstract:
We present an approach for state and parameter estimation in retinal laser treatment by a novel setup where both measurement and heating is performed by a single laser. In this medical application, the temperature that is induced by the laser in the patient's eye is critical for a successful and safe treatment. To this end, we pursue a model-based approach using a model given by a heat diffusion e…
▽ More
We present an approach for state and parameter estimation in retinal laser treatment by a novel setup where both measurement and heating is performed by a single laser. In this medical application, the temperature that is induced by the laser in the patient's eye is critical for a successful and safe treatment. To this end, we pursue a model-based approach using a model given by a heat diffusion equation on a cylindrical domain, where the source term is given by the absorbed laser power. The model is parametric in the sense that it involves an absorption coefficient, which depends on the treatment spot and plays a central role in the input-output behavior of the system. After discretization, we apply a particularly suited parametric model order reduction to ensure real-time tractability while retaining parameter dependence. We augment known state estimation techniques, i.e., extended Kalman filtering and moving horizon estimation, with parameter estimation to estimate the absorption coefficient and the current state of the system. Eventually, we show first results for simulated and experimental data from porcine eyes. We find that, regarding convergence speed, the moving horizon estimation slightly outperforms the extended Kalman filter on measurement data in terms of parameter and state estimation, however, on simulated data the results are very similar.
△ Less
Submitted 4 March, 2021;
originally announced March 2021.
