AN APPROACH TO SOLVE THE TRACKING PROBLEM

摘要

We show in this paper a method to reach the asymptotical tracking of partially unknown nonlinear systems with neural networks and the help of back-stepping alike procedures. Besides, within our result we will see some improvements we have accomplished to the work of [12] where they have analyzed systems without internal dynamics and have demonstrated the stability, but not asymptotic stability, of the tracking error, however they impose some bounds which happen to vanish the tracking error at expense of exciting the unmodeled dynamics. This work relies primarily in a change of coordinates that brings the system into one whose first coordinates are given underlying the method proposed in [12], and the last ones with a structure which can help us in finding a control law which guarantees the asymptotic stability of them as well as the asymptotic vanishing of the control law itself, all this will assure the asymptotic stability of a tracking error.