arXiv subjects

Frank Allgöwer

Publications and source records attributed to Frank Allgöwer.

At least 73 records · Page 4Linked to original sources

arXiv2024

Data-driven Dissipativity Analysis of Linear Parameter-Varying Systems

We derive direct data-driven dissipativity analysis methods for Linear Parameter-Varying (LPV) systems using a single sequence of input-scheduling-output data. By means of constructing a semi-definite program subject to linear matrix inequality constraints based on this data-dictionary, direct data-driven verification of $(Q,S,R)$-type of dissipativity properties of the data-generating LPV system is achieved. Multiple implementation methods are proposed to achieve efficient computational properties and to even exploit structural information on the scheduling, e.g., rate bounds. The effectiveness and trade-offs of the proposed methodologies are shown in simulation studies of academic and physically realistic examples.

eess.SY↗

arXiv2024

On $\ell_2$-performance of weakly-hard real-time control systems

This paper considers control systems with failures in the feedback channel, that occasionally lead to loss of the control input signal. A useful approach for modeling such failures is to consider window-based constraints on possible loss sequences, for example that at least r control attempts in every window of s are successful. A powerful framework to model such constraints are weakly-hard real-time constraints. Various approaches for stability analysis and the synthesis of stabilizing controllers for such systems have been presented in the past. However, existing results are mostly limited to asymptotic stability and rarely consider performance measures such as the resulting $\ell_2$-gain. To address this problem, we adapt a switched system description where the switching sequence is constrained by a graph that captures the loss information. We present an approach for $\ell_2$-performance analysis involving linear matrix inequalities (LMI). Further, leveraging a system lifting method, we propose an LMI-based approach for synthesizing state-feedback controllers with guaranteed $\ell_2$-performance. The results are illustrated by a numerical example.

eess.SY↗

arXiv2024

Distributed MPC for Self-Organized Cooperation of Multiagent Systems -- Extended Version

We present a sequential distributed model predictive control (MPC) scheme for cooperative control of multi-agent systems with dynamically decoupled heterogeneous nonlinear agents subject to individual constraints. In the scheme, we explore the idea of using tracking MPC with artificial references to let agents coordinate their cooperation without external guidance. Each agent combines a tracking MPC with artificial references, the latter penalized by a suitable coupling cost. They solve an individual optimization problem for this artificial reference and an input that tracks it, only communicating the former to its neighbors in a communication graph. This puts the cooperative problem on a different layer than the handling of the dynamics and constraints, loosening the connection between the two. We provide sufficient conditions on the formulation of the cooperative problem and the coupling cost for the closed-loop system to asymptotically achieve it. Since the dynamics and the cooperative problem are only loosely connected, classical results from distributed optimization can be used to this end. We illustrate the scheme's application to consensus and formation control.

eess.SY↗

arXiv2024

Collision Avoidance Safety Filter for an Autonomous E-Scooter using Ultrasonic Sensors

In this paper, we propose a collision avoidance safety filter for autonomous electric scooters to enable safe operation of such vehicles in pedestrian areas. In particular, we employ multiple low-cost ultrasonic sensors to detect a wide range of possible obstacles in front of the e-scooter. Based on possibly faulty distance measurements, we design a filter to mitigate measurement noise and missing values as well as a gain-scheduled controller to limit the velocity commanded to the e-scooter when required due to imminent collisions. The proposed controller structure is able to prevent collisions with unknown obstacles by deploying a reduced safe velocity ensuring a sufficiently large safety distance. The collision avoidance approach is designed such that it may be easily deployed in similar applications of general micromobility vehicles. The effectiveness of our proposed safety filter is demonstrated in real-world experiments.

eess.SY↗

arXiv2024

Koopman-based feedback design with stability guarantees

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.

eess.SY↗

arXiv2024

Time- versus event-triggered consensus of a single-integrator multi-agent system

Event-triggered control has shown the potential for providing improved control performance at the same average sampling rate when compared to time-triggered control. While this observation motivates numerous event-triggered control schemes, proving it from a theoretical perspective has only been achieved for a limited number of settings. Inspired by existing performance analyses for the single-loop case, we provide a first fundamental performance comparison of time- and event-triggered control in a multi-agent consensus setting. For this purpose, we consider undirected connected network topologies without communication delays, a level-triggering rule for event-triggered control, and the long-term average of the quadratic deviation from consensus as a performance measure. The main finding of our analysis is that time-triggered control provably outperforms event-triggered control beyond a certain number of agents in our particular setting. We thereby provide an illustrative distributed problem setup in which event-triggered control results in a performance disadvantage when compared to time-triggered control in the case of large networks. Moreover, we derive the asymptotic order of the performance measure under both triggering schemes which gives more insights into the cost relationship for large numbers of agents. Thus, by presenting an analysis for a particular setup, this work points out that transferring an event-triggering scheme from the single-loop to the multi-agent setting can lead to a loss of the often presumed superiority of event-triggered control over time-triggered control. In particular, the design of performant decentralized event-triggering schemes can therefore pose additional challenges when compared to the analogue single-loop case.

eess.SY↗

arXiv2024

Bootstrapping Guarantees: Stability and Performance Analysis for Dynamic Encrypted Control

Encrypted dynamic controllers that operate for an unlimited time have been a challenging subject of research. The fundamental difficulty is the accumulation of errors and scaling factors in the internal state during operation. Bootstrapping, a technique commonly employed in fully homomorphic cryptosystems, can be used to avoid overflows in the controller state but can potentially introduce significant numerical errors. In this paper, we analyze dynamic encrypted control with explicit consideration of bootstrapping. By recognizing the bootstrapping errors occurring in the controller's state as an uncertainty in the robust control framework, we can provide stability and performance guarantees for the whole encrypted control system. Further, the conservatism of the stability and performance test is reduced by using a lifted version of the control system.

eess.SY↗

arXiv2024

Lipschitz-bounded 1D convolutional neural networks using the Cayley transform and the controllability Gramian

We establish a layer-wise parameterization for 1D convolutional neural networks (CNNs) with built-in end-to-end robustness guarantees. In doing so, we use the Lipschitz constant of the input-output mapping characterized by a CNN as a robustness measure. We base our parameterization on the Cayley transform that parameterizes orthogonal matrices and the controllability Gramian of the state space representation of the convolutional layers. The proposed parameterization by design fulfills linear matrix inequalities that are sufficient for Lipschitz continuity of the CNN, which further enables unconstrained training of Lipschitz-bounded 1D CNNs. Finally, we train Lipschitz-bounded 1D CNNs for the classification of heart arrythmia data and show their improved robustness.

cs.LG↗

arXiv2024

Novel Quadratic Constraints for Extending LipSDP beyond Slope-Restricted Activations

Recently, semidefinite programming (SDP) techniques have shown great promise in providing accurate Lipschitz bounds for neural networks. Specifically, the LipSDP approach (Fazlyab et al., 2019) has received much attention and provides the least conservative Lipschitz upper bounds that can be computed with polynomial time guarantees. However, one main restriction of LipSDP is that its formulation requires the activation functions to be slope-restricted on $[0,1]$, preventing its further use for more general activation functions such as GroupSort, MaxMin, and Householder. One can rewrite MaxMin activations for example as residual ReLU networks. However, a direct application of LipSDP to the resultant residual ReLU networks is conservative and even fails in recovering the well-known fact that the MaxMin activation is 1-Lipschitz. Our paper bridges this gap and extends LipSDP beyond slope-restricted activation functions. To this end, we provide novel quadratic constraints for GroupSort, MaxMin, and Householder activations via leveraging their underlying properties such as sum preservation. Our proposed analysis is general and provides a unified approach for estimating $\ell_2$ and $\ell_\infty$ Lipschitz bounds for a rich class of neural network architectures, including non-residual and residual neural networks and implicit models, with GroupSort, MaxMin, and Householder activations. Finally, we illustrate the utility of our approach with a variety of experiments and show that our proposed SDPs generate less conservative Lipschitz bounds in comparison to existing approaches.

cs.LG↗

arXiv2024

Transient Performance of MPC for Tracking

We analyse the closed-loop performance of a model predictive control (MPC) for tracking formulation with artificial references. It has been shown that such a scheme guarantees closed-loop stability and recursive feasibility for any externally supplied reference, even if it is unreachable or time-varying. The basic idea is to consider an artificial reference as an additional decision variable and to formulate generalised terminal ingredients with respect to it. In addition, its offset is penalised in the MPC optimisation problem, leading to closed-loop convergence to the best reachable reference. In this paper, we provide a transient performance bound on the closed loop using MPC for tracking. We employ mild assumptions on the offset cost and scale it with the prediction horizon. In this case, an increasing horizon in MPC for tracking recovers the infinite horizon optimal solution.

math.OC↗

arXiv2024

Analysis and design of model predictive control frameworks for dynamic operation -- An overview

This article provides an overview of model predictive control (MPC) frameworks for dynamic operation of nonlinear constrained systems. Dynamic operation is often an integral part of the control objective, ranging from tracking of reference signals to the general economic operation of a plant under online changing time-varying operating conditions. We focus on the particular challenges that arise when dealing with such more general control goals and present methods that have emerged in the literature to address these issues. The goal of this article is to present an overview of the state-of-the-art techniques, providing a diverse toolkit to apply and further develop MPC formulations that can handle the challenges intrinsic to dynamic operation. We also critically assess the applicability of the different research directions, discussing limitations and opportunities for further research.

eess.SY↗

arXiv2024

Linearly discounted economic MPC without terminal conditions for periodic optimal operation

In this work, we study economic model predictive control (MPC) in situations where the optimal operating behavior is periodic. In such a setting, the performance of a standard economic MPC scheme without terminal conditions can generally be far from optimal even with arbitrarily long prediction horizons. Whereas there are modified economic MPC schemes that guarantee optimal performance, all of them are based on prior knowledge of the optimal period length or of the optimal periodic orbit itself. In contrast to these approaches, we propose to achieve optimality by multiplying the stage cost by a linear discount factor. This modification is not only easy to implement but also independent of any system- or cost-specific properties, making the scheme robust against online changes therein. Under standard dissipativity and controllability assumptions, we can prove that the resulting linearly discounted economic MPC without terminal conditions achieves optimal asymptotic average performance up to an error that vanishes with growing prediction horizons. Moreover, we can guarantee practical asymptotic stability of the optimal periodic orbit under the additional technical assumption that dissipativity holds with a continuous storage function. We complement these qualitative guarantees with a quantitative analysis of the transient and asymptotic average performance of the linearly discounted MPC scheme in a numerical simulation study.

eess.SY↗

arXiv2023

Guarantees for data-driven control of nonlinear systems using semidefinite programming: A survey

This survey presents recent research on determining control-theoretic properties and designing controllers with rigorous guarantees using semidefinite programming and for nonlinear systems for which no mathematical models but measured trajectories are available. Data-driven control techniques have been developed to circumvent a time-consuming modelling by first principles and because of the increasing availability of data. Recently, this research field has gained increased attention by the application of Willems' fundamental lemma, which provides a fertile ground for the development of data-driven control schemes with guarantees for linear time-invariant systems. While the fundamental lemma can be generalized to further system classes, there does not exist a comparable data-based system representation for nonlinear systems. At the same time, nonlinear systems constitute the majority of practical systems. Moreover, they include additional challenges such as data-based surrogate models that prevent system analysis and controller design by convex optimization. Therefore, a variety of data-driven control approaches has been developed with different required prior insights into the system to ensure a guaranteed inference. In this survey, we will discuss developments in the context of data-driven control for nonlinear systems. In particular, we will focus on methods based on system representations providing guarantees from finite data, while the analysis and the controller design boil down to convex optimization problems given as SDP. Thus, these approaches achieve reasonable advances compared to the state-of-the-art system analysis and controller design by models from system identification. Specifically, the paper covers system representations based on extensions of Willems' fundamental lemma, set membership, kernel techniques, the Koopman operator, and feedback linearization.

math.OC↗

arXiv2023

Data-driven system analysis of nonlinear systems using polynomial approximation

In the context of data-driven control of nonlinear systems, many approaches lack of rigorous guarantees, call for nonconvex optimization, or require knowledge of a function basis containing the system dynamics. To tackle these drawbacks, we establish a polynomial representation of nonlinear functions based on a polynomial sector by Taylor's theorem and a set-membership for Taylor polynomials. The latter is obtained from finite noisy samples. By incorporating the measurement noise, the error of polynomial approximation, and potentially given prior knowledge on the structure of the system dynamics, we achieve computationally tractable conditions by sum of squares relaxation to verify dissipativity and incremental dissipativity of nonlinear dynamical systems with rigorous guarantees. The framework is extended by combining multiple Taylor polynomial approximations which yields a less conservative piecewise polynomial system representation. The proposed approach is applied for numerical and experimental examples. There it is compared to a least-squares-error model including knowledge from first principle.

eess.SY↗

arXiv2023

Control of bilinear systems using gain-scheduling: Stability and performance guarantees

In this paper, we present a state-feedback controller design method for bilinear systems. To this end, we write the bilinear system as a linear fractional representation by interpreting the state in the bilinearity as a structured uncertainty. Based on that, we derive convex conditions in terms of linear matrix inequalities for the controller design, which are efficiently solvable by semidefinite programming. Further, we prove asymptotic stability and quadratic performance of the resulting closed-loop system locally in a predefined region. The proposed design uses gain-scheduling techniques and results in a state feedback with rational dependence on the state, which can substantially reduce conservatism and improve performance in comparison to a simpler, linear state feedback. Moreover, the design method is easily adaptable to various scenarios due to its modular formulation in the robust control framework. Finally, we apply the developed approaches to numerical examples and illustrate the benefits of the approach.

eess.SY↗

arXiv2023

Performance implications of different $p$-norms in level-triggered sampling

This work studies the performance of an event-based control approach, namely level-triggered sampling, in a standard multidimensional single-integrator setup. We falsify a conjecture from the literature that the deployed $p$-norm in the triggering condition supposedly has no impact on the performance of the sampling scheme in that setting. In particular, we show for the considered setup that the usage of the maximum norm instead of the Euclidean norm induces a performance deterioration of level-triggered sampling for sufficiently large system dimensions, when compared to periodic control at the same sampling rate. Moreover, we investigate the performance for other $p$-norms in simulation and observe that it degrades with increasing $p$. In addition, our findings reveal the previously unknown role of the triggering rule in the cause of a recently discovered phenomenon: Previous work has shown for a single-integrator consensus setup that the commonly observed performance advantage of event-based control over periodic control can be lost in distributed settings with a cooperative control goal. In our work, we obtain similar results for a non-cooperative setting only by adjusting the norm in the level-triggered sampling scheme. We therefore demonstrate that the performance degradation found in the distributed setting originates from the triggering rule and not from the considered cooperative control goal.

eess.SY↗

arXiv2023

Data-driven distributed MPC of dynamically coupled linear systems

In this paper, we present a data-driven distributed model predictive control (MPC) scheme to stabilise the origin of dynamically coupled discrete-time linear systems subject to decoupled input constraints. The local optimisation problems solved by the subsystems rely on a distributed adaptation of the Fundamental Lemma by Willems et al., allowing to parametrise system trajectories using only measured input-output data without explicit model knowledge. For the local predictions, the subsystems rely on communicated assumed trajectories of neighbours. Each subsystem guarantees a small deviation from these trajectories via a consistency constraint. We provide a theoretical analysis of the resulting non-iterative distributed MPC scheme, including proofs of recursive feasibility and (practical) stability. Finally, the approach is successfully applied to a numerical example.

eess.SY↗

arXiv2023

Koopman interpretation and analysis of a public-key cryptosystem: Diffie-Hellman key exchange

The security of public-key cryptosystems relies on computationally hard problems, that are classically analyzed by number theoretic methods. In this paper, we introduce a new perspective on cryptosystems by interpreting the Diffie-Hellman key exchange as a nonlinear dynamical system. Employing Koopman theory, we transfer this dynamical system into a higher-dimensional space to analytically derive a purely linear system that equivalently describes the underlying cryptosystem. In this form, analytic tools for linear systems allow us to reconstruct the secret integers of the key exchange by simple manipulations. Moreover, we provide an upper bound on the minimal required lifting dimension to obtain perfect accuracy. To demonstrate the potential of our method, we relate our findings to existing results on algorithmic complexity. Finally, we transfer this approach to a data-driven setting where the Koopman representation is learned from data samples of the cryptosystem.

eess.SY↗