Ahmad Poursaberi - Faculty of Computer and Electrical Engineering, University of Tehran, Iran
Mohammad Javd Yazdanpanah - Faculty of Computer and Electrical Engineering, University of Tehran, Iran

This paper proposes an adaptive backstepping design approach for a kind of system in strict-feedback form using multilayered neural network. An online adoption of the network is achieved without any offline training. The network also simplifies design of the backstepping by obviating need to build a regressor and basis functions. The architecture is so simple and easily develops for any number of steps in backstepping method. At last this approach is applied to a numerical example.

Backstepping Control, Neural Networks, Adaptive Control, Strict-Feedback Form, Nonlinear Systems