Paper Title
Adaptive Backstepping Control of Two-Link Positioning System Using RBF Neural Network For Estimation of Uncertainties

Abstract
Two-axis pan-tilt systems are widely used for positioning system. When it is installed to be operated on a moving platform like a vehicle, the motion of the platform affects the positioning performance significantly. In this paper, a radial basis function (RBF) network is used for estimating the external disturbances and the estimated value is used for compensation in the control law. Using the RBF neural networks (NN) we designed Adaptive Backstepping controller for estimating a disturbance compensation and error. In this process, we analyzed stability based on Lyapunov through uniformly ultimately bounded. Through computer simulations, we tested various disturbances in which frequency and magnitude varies irregularly. In RBF NN, minimized input vector was designed and estimation compensation process of the disturbance was analyzed. As a result, two input vectors were designed in RBF NN, and it improved the performance of the backstepping controller which means robustness is secured using the disturbance observer. Index Terms—RBF Neural Network, Backstepping, Disturbance Observer, 2-Axis Pan-tilt.