Designing a completely distributed observer with robustness against disturbances

Distributed observer is an effective scheme for state estimation of a targeted system, utilizing a network of sensors to perform local output measurements. In this article, the problem of designing a distributed observer for a linear time-invariant system with external disturbances is addressed. The objective is to perform state estimation of such a system while simultaneously mitigating the impact of external disturbances and maintaining the distributed communication characteristics among the sensors. To achieve this, a completely distributed observer with robustness against disturbances is developed by combining the high-gain observer technique and the adaptive coupling strength strategy. Initially, a traditional high-gain observer is used for the state estimation on the observable subsystem, and subsequently a high-gain observer without peaking is introduced to avoid the peaking phenomenon. Finally, numerical simulations are presented to verify the theoretical results. © 2024 John Wiley & Sons Ltd.