UDE-based Consensus Tracking Control of Multi-Agent Systems with Actuator Faults and External Disturbances

Abstract

The main objective of this paper is to address the robust consensus tracking control problem for multi-agent systems, despite the presence of actuator faults and external disturbances. By integrating the techniques of neural network (NN) and uncertainty and disturbance estimator (UDE), a distributed controller is formulated for each follower agent. At first, UDE is constructed to estimate the external disturbances of the follower agent. The estimation performance is adjusted by tuning a simple parameter through a trade-off between the ability of the controller and cooperative performance. Then, NN is utilized to estimate the leader’s uncertainties and generate a high-order reference signal for the followers. Meanwhile, the adaptive update laws which are designed for the parameter estimates of NN are formulated and executed online. Based on the Lyapunov stability theorem, asymptotically consensus tracking can be achieved with the undirected and connected communication topology. Finally, some numerical simulations are conducted to validate the effectiveness of the proposed control scheme, and the results are presented.