Distributed Adaptive Tracking Control for a Class of High-order Uncertain Nonlinear Multi-agent Systems via Neural Approximation Approach

This research aims at establishing a distribution adaptive tracking strategy for a class of high-order non-strict feedback uncertain nonlinear multi-agent systems. In each follower dynamics, the virtual and real control items are positive odd integer power functions instead of common linear items. In addition, each follower’s dynamics appears in non-strict feedback form. In controller design process, adding a power integrator approach is adopted to overcome the difficulty caused by high-order structural feature. Radial basis function neural networks are adopted to approximate the unknown nonlinearities. Further, by adaptive neural backstepping technique an adaptive tracking proposal is proposed to ensure that the signals in the whole closed-loop system are bounded and all the outputs of followers synchronously track the leader’s signal. A numerical simulation is included to further verify our results.