Robust neural net-based MRAC of a class of nonlinear plants with application to robotic control

Robust neural net-based MRAC of a class of nonlinear plants with application to robotic control

  In this paper, we discussed the development of robust neural net-based MRAC scheme for a class of MIMO nonlinear plants. It is assumed that the class of nonlinear plants could be described by a kind of pseudo-linear state space model. The plants need not to be linearizable in general. Hence, a robust neural net-based MRAC scheme is proposed for them, with VSC technique incorporated in it. Further, we have applied the proposed scheme to the control of an industrial robotic manipulator, to demonstrate how it can be used in the field of robotic control. Stability analysis has been performed by use of Lyapunov method. Simulation studies on the control of the industrial robotic manipulator are carried out to validate the proposed scheme.