A new operator-based framework for parameter-dependent Vlasov-Maxwell dynamics is presented as a way to show, under its assumptions, that the model has a unique mild solution, the closed-loop system can be uniformly exponentially stable, and an observer's estimation error can converge uniformly exponentially when the paper's stated inequalities hold. The authors also give an H-infinity condition for zero-disturbance exponential stability and for a specified disturbance-attenuation property when the initial condition is zero.
The claims are conditional on the model assumptions. Under Assumptions 1 to 4, the existence result says that each admissible parameter trajectory and input has a unique mild solution. Here, "mild solution" means the time-evolving state described through the model's evolution operators. The paper also states that the growth constants governing that evolution are uniform across the allowed parameter set, relying on a Kato stability assumption and compactness of the set.
From abstract conditions to workable tests
The central stability check is a primal operator differential linear matrix inequality, or LMI. In plain language, it is an inequality involving the model's operators and a matrix-valued quantity that can act as a stability certificate. If the condition is satisfied, the paper states that every closed-loop solution obeys a uniform exponential stability estimate. A dual operator differential LMI supplies the parallel observer result: it is presented as a sufficient condition for the observer's estimation error to converge uniformly and exponentially.
To turn those operator conditions into computations, the paper uses Galerkin projections, a way of reducing an operator problem to finite-dimensional coordinates. The authors state that the resulting finite-dimensional LMIs remain consistent with the underlying operator inequalities. This gives the framework a direct route from its abstract conditions to numerical controller and observer synthesis.
A deliberately small numerical demonstration
The numerical demonstration is deliberately small. It uses one spatial dimension, a single velocity variable, four Hermite modes and periodic boundaries. The truncation is six in space and four in velocity, producing a 48-dimensional model. The scheduling parameter was prescribed as a sine pattern centered at 1.1, with an amplitude of 0.25 and a time factor of 0.3. Its allowed range was 0.8 to 1.4, and its absolute rate of change was no more than 0.7.
Both the L2 observer design and the H-infinity design produced stable, well-conditioned simulations, but the selected solver returned the status 'optimal_inaccurate' in all experiments. The reported L2 beta value was 1.21 times 10 to the minus 5. For H-infinity, beta was 7.82 times 10 to the minus 6 and the optimized attenuation level, gamma, was 1.57 times 10 to the minus 8. The empirical attenuation was approximately 6.26, a substantial difference from the optimized figure.
Convergence came with a numerical warning
At time 20, the reported electric-field error was approximately 10 to the minus 3 and the kinetic error approximately 3 times 10 to the minus 2. Both designs had negative fitted asymptotic rates, consistent with declining errors in the late-time part of the reduced simulations. For L2, the electric-field rate was minus 2.39 times 10 to the minus 2 and the kinetic rate was minus 9.56 times 10 to the minus 2. For H-infinity, the corresponding rates were minus 5.43 times 10 to the minus 2 and minus 2.33 times 10 to the minus 2. The rates were estimated from the asymptotic simulation regime, and the benchmark supplied no statistical uncertainty estimates.
The boundaries of the result are as important as its headline numbers. The analysis assumes periodic spatial boundaries and excludes reflecting or absorbing ones. The numerical example is reduced and one-dimensional, so it does not demonstrate high-fidelity three-dimensional plasma control or observation, performance in physical-plasma experiments, or statistical generalization. It also does not establish that the optimized H-infinity bound is numerically accurate.
The paper is an arXiv version 1 preprint dated 28 August 2026. Funding information was not reported in the supplied text, while the authors reported no potential conflict of interest. Its contribution is best read as a conditional synthesis framework: the theory supplies sufficient conditions, and the small calculations show how those conditions can be implemented, but they leave the gap between optimized and empirical attenuation unresolved.
Paper data and sources
Original title: Operator-Theoretic Stability and Observer Synthesis for Parameter-Dependent Vlasov--Maxwell Dynamics
Authors: Amadou Cissé, Mohamed Boutayeb
Journal/Repository: arXiv
Status: Preprint, not yet peer-reviewed
First online: 2026-08-28
DOI: Not available
Original paper · Full text