Leader-Following Consensus of Multiple Euler-Lagrange Systems Under Deception Attacks

Abstract

In this paper, the leader-following consensus problem of a multi-Euler-Lagrange system is studied under deception attacks, where attackers can inject the false data into the data being exchanged over the communication network between two Euler-Lagrange systems. A distributed adaptive filter is proposed to compensate the unknown injection false data, where the magnitude of the false data is estimated. Meanwhile, the performance of the multiple Euler-Lagrange system consensus with the proposed filter can be guaranteed to be similar to that with the traditional controller when the communication network is not injected into false data by a malicious attacker. Furthermore, the consensus criteria of a multiple Euler-Lagrange system and the parameter design scheme of the proposed adaptive filter are achieved by using Lyapunov stability theory. Finally, a numerical example is carried out to demonstrate the effectiveness of the proposed consensus controller for a multi-Euler-Lagrange system attacked by using the false data injection.INDEX TERMS Euler-Lagrange systems, leader-following consensus, deception attack, distributed filter.

Resilient and Privacy-Preserving Leader-Follower Consensus in Presence of Cyber-Attacks

“…Since f (x 0 (t)), g(z 0 (t)), p i j,1 (t), p i j,2 (t), δ ui (t), δ ji (t), δ ji (t) are all uniformly bounded, the vector ∆(t) is also uniformly bounded. The solution to (15) is given by…”

“…While the individual agent in those work has linear or non-linear dynamics, the type of attacks is only limited to DoS attacks which are less sophisticated than the false data injection (FDI) attacks. On the other hand, the work [12], [13], [15] present approaches for resilient leader-follower consensus under FDI attacks on the communication links. However, the work do not consider attacks on the node's actuator.…”

Event-triggered consensus control for a class of uncertain multiple Euler-Lagrange systems with actuator faults

“…In fact, the nonlinear dynamic matrices must be partially known in [34]. Although all the nonlinear dynamic matrices are unknown in [27,13,16,5,24,8,20,26], external disturbance is excluded therein. In [28,17,11,9,2,19], the external disturbance is considered, but the dynamic matrices are required to be partially known in [17] or have known nominal parts in [19], while those in other literature must be parametrically separable.…”

“…2 ○ More measurements of the leader are required for control design. Specifically, the time derivatives of the leader's output in the first or second order are required to be available for feedback in [34,24,20,26,28,9,19,4,33]. This implies that the velocity or even the acceleration of the leader must be measurable.…”

“…Actuator faults will introduce extra nonlinearity and uncertainties, and hence bring new essential obstacles in the control design. However, most of the works [34,27,13,16,5,24,8,20,26,28,17,11,9,2,19,4,33] in this topic ignore the presence of actuator faults. Although few works consider actuator faults [3,21,32], the results are constrained ecause the time derivatives of the leader's outputs are required to be available for feedback in [3,21,32], while system parameters are partially known in [21], and the unknown system matrices must be parametrically separable in [21,32].…”