Neural-Adaptive Finite-Time Consensus Tracking of Singular Multiagent Systems With Actuator Attack.

Abstract

Since the actuator attack is an important threat to the complicated network, the problem of finite-time consensus tracking (FTCT) control for time-varying delay singular multiagent systems with actuator attack is investigated in this article. First, to approximate the states of the follower agents and design the sliding mode surfaces under actuator attack, a state observer is proposed. Then, the novel neural-adaptive distributed FTCT protocol and actuator attack estimation protocol are designed based on the radial basis function neural network estimation method. The sliding mode control with a partitioning strategy is adopted for the FTCT problem analysis of singular multiagent systems. The sufficient conditions and solving strategies of FTCT for singular multiagent systems over the whole finite-time interval are provided. Finally, the validity of FTCT for singular multiagent systems is verified by simulation examples.