Adaptive Neural Network Controller Design for Flexible Joint Robots using Singular Perturbation Technique

Abstract

In this paper, an adaptive neural network controller is presented for flexible joint robots using Singular Perturbation techniques by modelling the elastic forces as the fast variables and link variables as slow variables. The neural network controller is to control the slow dynamics in order to eliminate the tedious preliminary computation of the regressor matrix. Unlike many neural network (NN) controllers in the literature, inverse dynamical model evaluation is not required and no time-consuming training process is necessary except for initialising the NNs based on approximate function values at the initial posture at time t = 0. It can be shown that the controller can control the system successfully by intensive computer simulation tests.