Neural Network Nonlinearity Compensation for a Mobile Manipulator

Jun Oh Jang

Abstract: A control structure that makes possible the integration of a linear controller and a neural network (NN) compensator for a mobile manipulator is presented. The stability of the closed loop system and the boundedness of tracking errors are proved using Lyapunov theory. The NN compensation scheme proposed in this work can deal with unmodeled bounded nonlinearity and/or unstructured unmodeled dynamics in the mobile manipulator. On-line NN parameter tuning algorithms do no require off-line learning yet guarantee small tracking errors and bounded control signals are utilized.

Keywords: Mobile manipulator, Neural network, Compensation, Stability, Nonlinearity