Recursive KalmanNet: Analyse des capacit\'es de g\'en\'eralisation d'un r\'eseau de neurones r\'ecurrent guid\'e par un filtre de Kalman

TL;DR

This paper investigates the ability of Recursive KalmanNet, a neural network guided by a Kalman filter, to generalize to out-of-distribution scenarios with different temporal dynamics than training data.

Contribution

It provides an analysis of Recursive KalmanNet's generalization capabilities in scenarios with different temporal dynamics from training.

Findings

Recursive KalmanNet maintains estimation accuracy in out-of-distribution scenarios.

The model demonstrates robustness to changes in system dynamics.

Insights into the limits of neural network generalization in dynamic state estimation.

Abstract

The Recursive KalmanNet, recently introduced by the authors, is a recurrent neural network guided by a Kalman filter, capable of estimating the state variables and error covariance of stochastic dynamic systems from noisy measurements, without prior knowledge of the noise characteristics. This paper explores its generalization capabilities in out-of-distribution scenarios, where the temporal dynamics of the test measurements differ from those encountered during training. Le Recursive KalmanNet, r\'ecemment introduit par les auteurs, est un r\'eseau de neurones r\'ecurrent guid\'e par un filtre de Kalman, capable d'estimer les variables d'\'etat et la covariance des erreurs des syst\`emes dynamiques stochastiques \`a partir de mesures bruit\'ees, sans connaissance pr\'ealable des caract\'eristiques des bruits. Cet article explore ses capacit\'es de g\'en\'eralisation dans des sc\'enarios hors distribution, o\`u les dynamiques temporelles des mesures de test diff\`erent de celles rencontr\'ees \`a l'entra\^inement.