$\mathcal{H}_2/\mathcal{H}_{-}$ Distributed Fault Detection and Isolation for Heterogeneous Multi-Agent Systems

TL;DR

This paper presents a novel distributed fault detection and isolation method for heterogeneous multi-agent systems using $nd ased observer design, enabling agents to estimate states and detect faults robustly.

Contribution

It introduces a new $nd ased distributed observer approach for fault detection in heterogeneous multi-agent systems, ensuring robustness and sensitivity to faults.

Findings

The method effectively detects and isolates faults in multi-agent systems.

Agents can estimate their own and neighbors' states despite disturbances.

Numerical simulations confirm the approach's robustness and accuracy.

Abstract

The paper deals with the problem of distributed fault detection and isolation (FDI) for a group of heterogeneous multi-agent systems. The developed formation for the FDI is taken into account as a distributed observer design methodology, where the interaction between the agent and its neighbors is described as a vector of distributed relative output measurements. Based on two performance indexes $\mathcal{H}_2$ and $\mathcal{H}_{-}$, sufficient conditions are given to ensure the residual signals robust to the disturbances and sensitive with respect to the fault signals. In addition, we show that by using our proposed approach, each agent is able to estimate both its own states and states of its nearest neighbors in the presence of disturbances and faults. Finally, numerical simulations are provided to demonstrate the effectiveness of the theoretically analyzed results.