Event‐triggered distributed H∞$$ {H}_{\infty } $$ secure control for nonholonomic agents with dead‐zone inputs under attacks on sensors and actuators

Summary Distributed optimal control has been extensively investigated for nonholonomic mechanical agents (NMA). However, the existing methods have not yet taken into account vulnerabilities caused by malicious attacks on sensors and dead‐zone actuator links. The aim of this paper is therefore to pro...

Celý popis

Uloženo v:
Podrobná bibliografie
Vydáno v:International journal of robust and nonlinear control Ročník 34; číslo 2; s. 1238 - 1256
Hlavní autor: Tan, Luy Nguyen
Médium: Journal Article
Jazyk:angličtina
Vydáno: Bognor Regis Wiley Subscription Services, Inc 25.01.2024
Témata:
Actuators
Adaptive control
adaptive dynamic programming
cyber‐attacks
differential game
Differential games
Dynamic programming
event‐triggered control
Game theory
H-infinity control
observers
Optimal control
Robust control
Sensors
ISSN:1049-8923, 1099-1239
On-line přístup:Získat plný text
Tagy: Přidat tag
Žádné tagy, Buďte první, kdo vytvoří štítek k tomuto záznamu!
Popis
Shrnutí:Summary Distributed optimal control has been extensively investigated for nonholonomic mechanical agents (NMA). However, the existing methods have not yet taken into account vulnerabilities caused by malicious attacks on sensors and dead‐zone actuator links. The aim of this paper is therefore to propose a novel control scheme that not only achieves bounded L2$$ {L}_2 $$‐gain consensus performance and rejects dead‐zone disturbance but also mitigates the burden of communication resources and the effects of sensor and actuator attacks. Firstly, to estimate unmeasurable states and the unknown model of the attacks, event‐triggered (ET) observers are designed. Secondly, ET‐augmented control is proposed to transform Euler‐Lagrange dynamics into consensus tracking dynamics, from which the ET‐robust optimal control problem is formulated. Thirdly, the ET‐distributed H∞$$ {H}_{\infty } $$ secure control strategies are approximated synchronously via adaptive dynamic programming (ADP) and multi‐player differential game theory. Finally, a numerical example illustrates, on a networked multi‐robot system, the effectiveness of the proposed method.
Bibliografie:ObjectType-Article-1
SourceType-Scholarly Journals-1
ObjectType-Feature-2
content type line 14
ISSN:1049-8923
1099-1239
DOI:10.1002/rnc.7024