Adaptive NN formation tracking control for the multiple underactuated USVs with prescribed performance and input saturations

A novel neural network adaptive prescribed performance formation control algorithm is investigated for multiple underactuated unmanned surface vehicles (USVs) with asymmetric input saturations and unknown external disturbances. Initially, a new prescribed performance function is adopted to ensure that USV formation errors reach and keep within a predetermined range at a preset time. Then, the underactuated problem of USVs is addressed by an additional term generated by a Nussbaum function, and the model uncertain dynamics set is compensated by the radial basis function neural network (RBF NN) minimum parameter method. Meanwhile, adaptive techniques are employed to deal with external disturbances and unknown approximation errors. Furthermore, a continuously differentiable asymmetric saturation model based on the Gaussian error function is introduced to design the controller, eliminating the influence of non-smooth input saturations. Finally, the boundedness of all signals in the closed-loop system is given by the Lyapunov stability theory proof, and contrast simulation experiments verify the effectiveness and superiority of the proposed algorithm.

Automated summary

This paper investigates a novel neural network adaptive prescribed performance formation control algorithm for multiple underactuated unmanned surface vehicles (USVs) with asymmetric input saturations and unknown external disturbances. The algorithm addresses the underactuated problem of USVs and compensates the model uncertain dynamics using the radial basis function neural network (RBF NN) minimum parameter method. Adaptive techniques are employed to deal with external disturbances and unknown approximation errors. Furthermore, a continuously differentiable asymmetric saturation model is introduced to design the controller, eliminating the influence of non-smooth input saturations. The effectiveness and superiority of the proposed algorithm are verified through contrast simulation experiments.