ADAPTIVE ROBUST NEURAL NETWORK CONTROLLER BASIS ON BACKSTEPPING FOR ROBOT MANIPULATORS
Abstract
This present study proposes an Adaptive Robust Neural Networks
(ARNNs) based on backstepping control method for industrial robot
manipulators (IRMs) in order to improve high correctness of the position
tracking control. In this research, the ARNNs controller is combined the
advantages of Radial Basis Function Neural Networks (RBFNNs), the robust
term and adaptive backstepping control technique. The NNs is used in order to
approximate the unknown function to deal with external disturbances and
uncertain nonlinearities. In addition, the disturbance of system is
compensated by the robust Sliding Mode Control (SMC). All the parameters of
ARNNs are determined by the Lyapunov stability theorem. They are tuned
online by the adaptive training laws. Therefore, the stability, robustness and
desired tracking performance of ARNNs for IRMs are guaranteed. The
robustness and effectiveness of the ARNNs are proved by the simulations
performed on the three-link IRMs