Neural network adaptive dynamic sliding mode formation control of multi-agent systems

Abstract

This paper considers the problem of achieving time-varying formation for second-order multi-agent systems with actuator hysteresis, unknown system dynamics and external disturbances. A novel adaptive dynamic sliding mode scheme is developed to control a group of agents to follow desired trajectories. First, a dynamic sliding mode approach based on local formation tracking error is utilised to reject external disturbances and obtain smooth and chattering-free control input. Then, Chebyshev neural network is employed to estimate the nonlinear function related to the system's dynamic equation. A smooth projection law is also applied to regulate the output of the neural network. Moreover, a Bouc–Wen hysteresis compensator has been added to the current control law to offset the known actuator hysteresis effect. Finally, a numerical simulation based on a multiple omni-directional robot system is presented to illustrate the performance of the proposed control law.