Improving trajectory tracking of a three axis SCARA robot using neural networks

In this paper, a neural-network based robust adaptive controller is proposed to control an industrial robot considering non- linearities, uncertainties and external perturbations. Three-axis SCARA robots is used to test the performance of this controller. The nonlinear system is treated as a partially known system. The known dynamic is used to design a nominal feedback controller based on the well-known feedback linearization method and PD controller. A Variable Structure Controller is added to the PD loop to provide robustness to uncertainties in the model of the system in order to improve accuracy of the trajectory tracking. A Neural Network (NN) based robust adaptive tracking controller is applied to further improves the control action. The outputs of the NNs are used to compensate the effects of the system uncertainties and to improve the tracking performance. Using this scheme, strong robustness with respect to uncertain dynamics and nonlinearities can be obtained, the output tracking error between the plant output and the desired reference output can asymptotically converge to zero as well. This controller exhibited superior performance characteristics where the maximum absolute error for the three-axis SCARA robot is considerably reduced.