Cooperative Deterministic Learning-Based Trajectory Tracking for a Group of Unicycle-Type Vehicles

A cooperative deterministic learning based state feedback control algorithm is proposed in this paper for joint tracking control and learning/identification for a group of identical nonholonomic vehicles. Specifically, this algorithm is able to model the unknown nonlinear dynamics of the nonholonomic vehicle, and use it for trajectory tracking control with cooperative deterministic learning (DL) theory. In addition, cooperative DL grants every vehicle in the system the ability of knowledge learning not only along the trajectory of its own, but also along the trajectories of all other vehicles as well. It is shown using Lyapunov stability theory that with cooperative DL, the closed-loop system is guaranteed to be stable, with all vehicles tracking its own reference trajectories, and the radial basis function (RBF) neural network (NN) weights of all agents converge to the same constants.