Evan Arsenault | Searcharxiv

arXiv2023

Stability and long-time behaviour of a rigid body containing a damper

This paper provides a comprehensive analysis of stability and long-time behaviour of a coupled system constituted by two rigid bodies separated by a thin layer of lubricant. We show that permanent rotations of the whole system, with the solids at relative rest, are exponentially stable if and only if the axis of rotation is the principal axis of inertia corresponding to the largest moment of inertia of the outer body. All other equilibria are normally hyperbolic, and hence unstable. In addition, we show that all solutions to the governing equations converge to an equilibrium configuration, no matter which initial conditions are chosen. Numerical evidence of the above results as well as conditions ensuring attainability of the stable configurations are also presented. For the stability analysis we use a linearization principle for dynamical systems possessing a center manifold. The characterization of the long-time behaviour is obtained by a careful analysis of the partially dissipative system of equations governing the motion of the coupled system.

math.DS↗

arXiv2022

Toward Scalable Risk Analysis for Stochastic Systems Using Extreme Value Theory

We aim to analyze the behaviour of a finite-time stochastic system, whose model is not available, in the context of more rare and harmful outcomes. Standard estimators are not effective in making predictions about such outcomes due to their rarity. Instead, we use Extreme Value Theory (EVT), the theory of the long-term behaviour of normalized maxima of random variables. We quantify risk using the upper-semideviation $ρ(Y) = E(\max\{Y - μ,0\})$ of an integrable random variable $Y$ with mean $μ= E(Y)$. $ρ(Y)$ is the risk-aware part of the common mean-upper-semideviation functional $μ+ λρ(Y)$ with $λ\in [0,1]$. To assess more rare and harmful outcomes, we propose an EVT-based estimator for $ρ(Y)$ in a given fraction of the worst cases. We show that our estimator enjoys a closed-form representation in terms of the popular conditional value-at-risk functional. In experiments, we illustrate the extrapolation power of our estimator using a small number of i.i.d. samples ($<$50). Our approach is useful for estimating the risk of finite-time systems when models are inaccessible and data collection is expensive. The numerical complexity does not grow with the size of the state space.

stat.ME↗