Distributed Consensus Control of Networked Robotic Systems With Dynamic Leader Under Time-Varying Delays and Denial-of-Service Attacks

Abstract

Consensus of multi-agent systems has a wide range of applications, such as coordination, cooperative transportation, and synchronization. This study investigates consensus control in the task space of multi-robotic systems modeled using Euler-Lagrange equations with a moving leading under timevarying delays and denial-of-service (DoS) attacks, respectively. Two control algorithms are studied when networked robotic systems are interconnected over directed spanning trees with a constant-velocity leader as the root. First, we use the Lyapunov-Razumikhin theorem to address a condition of the maximum time delay for nonuniform time-delay communication, thereby ensuring asymptotic tracking of the network. Second, we propose a novel resilient control framework to deal with DoS attacks. Note that designing the control algorithm for DoS attacks is difficult because the attacks cause signal discontinuity. The proposed resilient controller can release the DoS duration condition, which is a common requirement in many related works. It is verified that the consensus errors converge to the origin under the proposed control frameworks. We present experimental results to demonstrate the effectiveness of the proposed controllers.INDEX TERMS Networked Euler-Lagrange systems, dynamic leader, time delays, DoS attacks, resilient control.