Event-triggered observer design for continuous-time linear networked systems

Abstract

In this paper, a novel event-triggered observer is proposed to decrease the information exchange in continuous-time linear networked systems. The structure of the proposed observer consists of two main parts. The first one, called the event-triggered mechanism (ETM), is placed by the system and is used to determine when the measured output of the system must be sent through the network. The ETM consists of a traditional Luenberger observer, an event-triggered local observer, and an event-detector. The second part is a copy of the event-triggered local observer which is used to estimate the system state using the transmitted data from the sensors. Under the observability assumption, it is proved that the error between the system state and its estimation is asymptotically limited to a predetermined bound. Moreover, it is guaranteed that the Zeno behaviour is avoided. Numerical simulations are provided to demonstrate the design procedure of the proposed event-triggered observer.