Kalman-like Observer for Hybrid Systems with Linear Maps and Known Jump Times (Full Version) - Archive ouverte HAL Accéder directement au contenu
Communication Dans Un Congrès Année : 2023

Kalman-like Observer for Hybrid Systems with Linear Maps and Known Jump Times (Full Version)

Gia Quoc Bao Tran
Pauline Bernard

Résumé

We propose a hybrid Kalman-like observer for general hybrid systems with linear (time-varying) dynamics and output maps, where the solutions' jump times are exactly known. After defining a hybrid observability Gramian and the corresponding hybrid uniform complete observability, we show that the estimate provided by this observer converges asymptotically to the system solution if this observability holds together with some boundedness and invertibility conditions along the considered system solution. Then, under additional uniformity and strictness of the forgetting factors, we show exponential stability of the estimation error with an arbitrarily fast rate. The robust stability of this error against input disturbances and measurement noise is also studied. The results are illustrated on several benchmark examples, including switched systems, hybrid systems with discontinuous solutions, and continuoustime systems with multi-rate sporadic outputs.

Domaines

Automatique
Fichier principal
Vignette du fichier
2023_IEEE CDC_Kalman.pdf (1.47 Mo) Télécharger le fichier
Origine Fichiers produits par l'(les) auteur(s)

Dates et versions

hal-04196352 , version 1 (05-09-2023)
hal-04196352 , version 2 (14-12-2023)

Identifiants

  • HAL Id : hal-04196352 , version 1

Citer

Gia Quoc Bao Tran, Pauline Bernard. Kalman-like Observer for Hybrid Systems with Linear Maps and Known Jump Times (Full Version). 62nd IEEE Conference on Decision and Control, Dec 2023, Singapore, Singapore. ⟨hal-04196352v1⟩
317 Consultations
136 Téléchargements

Partager

Gmail Mastodon Facebook X LinkedIn More