Fault-Tolerant Containment Control for Linear Multi-Agent Systems: An Adaptive Output Regulation Approach

Jie Zhang,Cuijuan An,Da-Wei Ding

doi:10.1109/access.2019.2926619

Abstract

This paper addresses the problem of fault-tolerant containment control (FTCC) for linear multi-agent systems (MASs) with process faults. First, distributed observers for followers are designed by utilizing their relative output estimation errors. Then, a new distributed adaptive FTCC law based upon the output regulation theory is proposed, where a distributed adaptive observer is implemented to estimate the synthesized information for leaders. It is shown that no matter whether there exist faults or not, all the followers can asymptotically move into a convex hull formed by the leaders. Finally, an example is given to illustrate the effectiveness of the designed distributed control law.

Highlights

Containment control can be seen as a consensus problem in multi-agent systems (MASs) with multiple leaders, aiming at driving all followers to a smallest convex region shaped by multiple leaders, e.g., a safe zone for working followers
When some process faults occur in follower agents i, i ∈ Vf, the fault-tolerant containment control (FTCC) problem is solved under a distributed FTCC law ui, i ∈ Vf such that follower-outputs yi, i ∈ Vf will move into the convex hull formed by leader-outputs wk, k ∈ Vl as time goes to infinity, i.e., lim t →∞
MAIN RESULTS a new distributed adaptive FTCC law is designed based upon the output regulation theory

Summary

Introduction

Containment control can be seen as a consensus problem in MASs with multiple leaders, aiming at driving all followers to a smallest convex region shaped by multiple leaders, e.g., a safe zone for working followers. Ye et al [18] solved the containment control problem of MASs with multiple leaders and actuator faults by designing a fault estimation observer. Different from the existing results [18], [23], [24], an adaptive output regulation approach is proposed to solve the FTCC problem of MASs with multiple leaders and process faults.

Results

Conclusion