Finite-frequency [formula omitted] unknown input observer-based distributed fault detection for multi-agent systems

Abstract

This paper focuses on fault detection and isolation (FDI) for a network of discrete-time multi-agent systems (MASs) with disturbances, in which a bank of finite-frequency H−/H∞ unknown input observers (UIOs) are constructed. To detect the fault in one agent, the unknown input consisting of faults in the other agents and disturbances in the whole system can be partitioned into two parts, the decoupled part and the non-decoupled part, which avoids the constraint of the strict rank condition of UIO. To make the generated residual sensitive to the corresponding fault while robust against the non-decoupled unknown input, a finite-frequency H−/H∞ UIO is proposed, of which the design conditions are transformed into a set of linear matrix inequalities (LMIs). A numerical simulation is conducted to illustrate the effectiveness of the proposed method.