Direct Adaptive Backstepping Control for a Class of MIMO Non-Affine Systems Using Recurrent Neural Networks

Free registration required

Executive Summary

In this paper, the backstepping technique is used to design adaptive controller for a class of MIMO nonlinear uncertain non-affine systems. The backstepping technique provides a systematic framework and recursive design methodology for nonlinear systems. The design procedure treats the state variables as virtual control inputs; then, the virtual controllers is designed step by step. Finally, the actual control input can be obtained. It illustrates the stability by Lyapunov stability theorem. However, the major constraint is that the system functions must be exactly known. If the internal uncertainty and external disturbance exist, then they may result in an unstable system. Therefore, the authors propose an Output-Recurrent Wavelet Neural Network (ORWNN) system to approximate the unknown functions to solve this problem.

  • Format: PDF
  • Size: 405.5 KB