arXiv subjects

Kirsten Morris

Publications and source records attributed to Kirsten Morris.

16 recordsLinked to original sources

arXiv2026

Verifiable Error Bounds for Physics-Informed Neural KKL Observers

This paper proposes a computable state-estimation error bound for learning-based Kazantzis--Kravaris/Luenberger (KKL) observers. Recent work learns the KKL transformation map with a physics-informed neural network (PINN) and a corresponding left-inverse map with a conventional neural network. However, no computable state-estimation error bounds are currently available for this approach. We derive a state-estimation error bound that depends only on quantities that can be certified over a prescribed region using neural network verification. We further extend the result to bounded additive measurement noise and demonstrate the guarantees on nonlinear benchmark systems.

eess.SY↗

arXiv2025

Stability of Jordan Recurrent Neural Network Estimator

State estimation refers to determining the states of a dynamical system that starts from a noisy initial condition and evolves under process noise, based on noisy measurements and a known system model. For linear dynamical systems with white Gaussian noises of known mean and variance, Kalman filtering is a well-known method that leads to stable error dynamics for detectable systems. There are some non-optimal extensions to nonlinear systems. Recent work has used neural networks to develop estimators for nonlinear systems that optimize a criterion. Stability of the error dynamics is even more important than optimality. Jordan recurrent neural networks (JRNs) have a structure that mimics that of a dynamical system and are thus appealing for estimator design. We show that a JRN performs better than an extended Kalman filter(EKF) and unscented Kalman filter(UKF) for several examples. The main contribution of this paper is an input-to-state stability analysis of the error dynamics of JRNs. The stability of the error dynamics of several examples is shown.

math.OC↗

arXiv2025

A Mixed-FEM approximation with uniform conservation of the exponential stability for a class of anisotropic port-Hamiltonian system and its application to LQ control

In this manuscript, we present a mixed finite element discretization for a class of boundary-damped anisotropic port-Hamiltonian systems. Using a multiplier method, we demonstrate that the resulting approximation model uniformly preserves the exponential stability of the uncontrolled system, establishing a lower bound for the exponential decay rate that is independent of the mesh size. This property is illustrated through the spatial discretization of a piezoelectric beam. Furthermore, we show how the uniform preservation of exponential stability by the proposed model aids in the convergence of controllers derived from an infinite-time linear quadratic control design, in comparison to models obtained from the standard finite-element method.

math.NA↗

arXiv2025

State estimator design using Jordan based long short-term memory networks

State estimation of a dynamical system refers to estimating the state of a system given an imperfect model, noisy measurements and some or no information about the initial state. While Kalman filtering is optimal for estimation of linear systems with Gaussian noises, calculation of optimal estimators for nonlinear systems is challenging. We focus on establishing a pathway to optimal estimation of high-order systems by using recurrent connections motivated by Jordan recurrent neural networks(JRNs). The results are compared to the corresponding Elman structure based long short-term memory network(ELSTM) and the KF for linear and EKF for nonlinear systems. The results suggest that for nonlinear systems, the use of long short-term memory networks can improve estimation error and also computation time. Also, the Jordan based long short-term memory networks(JLSTMs) require less training to achieve performance similar to ELSTMs.

math.OC↗

arXiv2024

On the Weierstraß form of infinite dimensional differential algebraic equations

The solvability for infinite dimensional differential algebraic equations possessing a resolvent index and a Weierstraß form is studied. In particular, the concept of integrated semigroups is used to determine a subset on which solutions exist and are unique. This information is later used for a important class of systems, namely, port-Hamiltonian differential algebraic equations.

math.AP↗

arXiv2024

Finite-time Linear-Quadratic Optimal Control of Partial Differential-Algebraic Equations

Finite-time linear-quadratic control of partial differential-algebraic equations (PDAEs) is considered. The discussion is restricted to those that are radial with index $0$; this corresponds to a nilpotency degree of 1. We establish the existence of a unique minimizing optimal control. A projection is used to derive a system of differential Riccati-like equation coupled with an algebraic equation, yielding the solution of the optimization problem in a feedback form. This generalizes the well-known result for PDEs to this class of PDAEs. These equations and hence the optimal control can be calculated without construction of the projected PDAE. Finally, we provide numerical simulations to illustrate application of the theoretical results.

math.OC↗

arXiv2024

Index concepts for linear differential-algebraic equations in finite and infinite dimensions

Different index concepts for linear differential-algebraic equations are defined in the general Banach space setting, and compared. For regular finite-dimensional linear differential-algebraic equations, all these indices exist and are equivalent. For infinite-dimensional systems, the situation is more complex. It is proven that although some indices imply others, in general they are not equivalent. The situation is illustrated with a number of examples.

math.DS↗

arXiv2023

Investigating Parameter Trainability in the SNAP-Displacement Protocol of a Qudit system

In this study, we explore the universality of Selective Number-dependent Arbitrary Phase (SNAP) and Displacement gates for quantum control in qudit-based systems. However, optimizing the parameters of these gates poses a challenging task. Our main focus is to investigate the sensitivity of training any of the SNAP parameters in the SNAP-Displacement protocol. We analyze conditions that could potentially lead to the Barren Plateau problem in a qudit system and draw comparisons with multi-qubit systems. The parameterized ansatz we consider consists of blocks, where each block is composed of hardware operations, namely SNAP and Displacement gates \cite{fosel2020efficient}. Applying Variational Quantum Algorithm (VQA) with observable and gate cost functions, we utilize techniques similar to those in \cite{mcclean2018barren} and \cite{cerezo2021cost} along with the concept of $t-$design. Through this analysis, we make the following key observations: (a) The trainability of a SNAP-parameter does not exhibit a preference for any particular direction within our cost function landscape, (b) By leveraging the first and second moments properties of Haar measures, we establish new lemmas concerning the expectation of certain polynomial functions, and (c) utilizing these new lemmas, we identify a general condition that indicates an expected trainability advantage in a qudit system when compared to multi-qubit systems.

quant-ph↗

arXiv2023

Boundary Control and Observer Design Via Backstepping for a Coupled Parabolic-Elliptic System

Stabilization of a coupled system consisting of a parabolic partial differential equation and an elliptic partial differential equation is considered. Even in the situation when the parabolic equation is exponentially stable on its own, the coupling between the two equations can cause instability in the overall system. A backstepping approach is used to derive a boundary control input that stabilizes the coupled system. The result is an explicit expression for the stabilizing control law. The second part of the paper involves the design of exponentially convergent observers to estimate the state of the coupled system, given some partial boundary measurements. The observation error system is shown to be exponentially stable, again by employing a backstepping method. This leads to the design of observer gains in closed-form. Finally, we address the output-feedback problem by combining the observers with the state feedback boundary control. The theoretical results are demonstrated with numerical simulations.

math.OC↗

arXiv2023

Exponential Decay Rate of Linear Port-Hamiltonian Systems. A Multiplier Approach

In this work, the multiplier method is extended to obtain a general lower bound of the exponential decay rate in terms of the physical parameters for port-Hamiltonian systems in one space dimension with boundary dissipation. The physical parameters of the system may be spatially varying. It is shown that under assumptions of boundary or internal dissipation, the system is exponentially stable. This is established through a Lyapunov function defined through a general multiplier function. Furthermore, an explicit bound on the decay rate in terms of the physical parameters is obtained. The method is applied to a number of examples.

math.AP↗

arXiv2022

On solvability of dissipative partial differential-algebraic equations

In this article we investigate the solvability of infinite-dimensional differential algebraic equations. Such equations often arise as partial differential-algebraic equations (PDAEs). A decomposition of the state-space that leads to an extension of the Hille-Yosida Theorem on Hilbert spaces for these equations is described. For dissipative partial differential equations the famous Lumer-Phillips generation theorem characterizes solvability and also boundedness of the associated semigroup. An extension of the Lumer-Phillips generation theorem to dissipative differential-algebraic equations is given. The results is illustrated by coupled systems and the Dzektser equation.

math.FA↗

arXiv2017

Efficient electrochemical model for lithium-ion cells

Lithium-ion batteries are used to store energy in electric vehicles. Physical models based on electro-chemistry accurately predict the cell dynamics, in particular the state of charge. However, these models are nonlinear partial differential equations coupled to algebraic equations, and they are computationally intensive. Furthermore, a variable solid-state diffusivity model is recommended for cells with a lithium ion phosphate positive electrode to provide more accuracy. This variable structure adds more complexities to the model. However, a low-order model is required to represent the lithium-ion cells' dynamics for real-time applications. In this paper, a simplification of the electrochemical equations with variable solid-state diffusivity that preserves the key cells' dynamics is derived. The simplified model is transformed into a numerically efficient fully dynamical form. It is proved that the simplified model is well-posed and can be approximated by a low-order finite-dimensional model. Simulations are very quick and show good agreement with experimental data.

math.AP↗

arXiv2016

Convergence Rate for the Ordered Upwind Method

The Ordered Upwind Method (OUM) is used to approximate the viscosity solution of the static Hamilton-Jacobi-Bellman (HJB) with direction-dependent weights on unstructured meshes. The method has been previously shown to provide a solution that converges to the exact solution, but no convergence rate has been theoretically proven. In this paper, it is shown that the solutions produced by the OUM in the boundary value formulation converge at a rate of at least the square root of the largest edge length in the mesh in terms of maximum error. An example with similar order of numerical convergence is provided.

math.OC↗

arXiv2015

Linearized stability analysis of nonlinear partial differential equations

Lyapunov's indirect method is an attractive method for analyzing stability of non-linear systems since only the stability of the corresponding linearized system needs to be determined. Unfortunately, the proof for finite-dimensional systems does not generalize to infinite-dimensions. In this paper a unified approach to Lyapunov's indirect method for infinite-dimensional system is described. It is shown how existing sufficient conditions fit this framework and a new sufficient condition is presented.

math.AP↗

arXiv2014

Root locii for systems defined on Hilbert spaces

The root locus is an important tool for analysing the stability and time constants of linear finite-dimensional systems as a parameter, often the gain, is varied. However, many systems are modelled by partial differential equations or delay equations. These systems evolve on an infinite-dimensional space and their transfer functions are not rational. In this paper a rigorous definition of the root locus for infinite-dimensional systems is given and it is shown that the root locus is well-defined for a large class of infinite-dimensional systems. As for finite-dimensional systems, any limit point of a branch of the root locus is a zero. However, the asymptotic behaviour can be quite different from that for finite-dimensional systems. This point is illustrated with a number of examples. It is shown that the familiar pole-zero interlacing property for collocated systems with a Hermitian state matrix extends to infinite-dimensional systems with self-adjoint generator. This interlacing property is also shown to hold for collocated systems with a skew-adjoint generator.

math.AP↗

arXiv2014

$C_0$-semigroups for hyperbolic partial differential equations on a one-dimensional spatial domain

Hyperbolic partial differential equations on a one-dimensional spatial domain are studied. This class of systems includes models of beams and waves as well as the transport equation and networks of non-homogeneous transmission lines. The main result of this paper is a simple test for $C_0$-semigroup generation in terms of the boundary conditions. The result is illustrated with several examples.

math.AP↗