State Estimation of Linear Systems over a Network subject to Sporadic Measurements, Delays, and Clock Mismatches

Marcello Guarro (1) , Francesco Ferrante (2) , Ricardo Sanfelice (1)

1 UCSC Department of Computer Engineering - Computer Engineering Department
2 GIPSA-SYSCO - GIPSA - Systèmes non linéaires et complexité

This paper proposes a hybrid observer for the estimation of the state of a plant over a realistic network. The network provides delayed measurements of the output of the plant at time instants that are not necessarily periodic and are accompanied by timestamps provided by a clock that eventually synchronizes with the clock of the observer. The proposed observer, along with the plant and communication network, are modeled by a hybrid dynamical system that has two timers, a logic variable, and two memory states to capture the mechanisms involved in the events associated with sampling and arrival of information, as well as the logic in the estimation algorithm. The hybrid model also includes a generic clock synchronization scheme to cope with a mismatch between the clocks at the plant and the observer. Convergence properties of the estimation error of the system are shown analytically and supported by numerical examples.

Domaines: Automatique

Marcello Guarro, Francesco Ferrante, Ricardo Sanfelice. State Estimation of Linear Systems over a Network subject to Sporadic Measurements, Delays, and Clock Mismatches. NecSys 2018 - 7th IFAC Workshop on Distributed Estimation and Control in Networked Systems, Aug 2018, Groningen, Netherlands. pp.313-318, ⟨10.1016/j.ifacol.2018.12.054⟩. ⟨hal-01850400⟩