NEURAL NETWORK BASED OBSERVER DESIGN FOR A CLASS OF UNKNOWN NONLINEAR SYSTEMS.

摘要

In this paper a new technique is developed to design an observer for unknown nonlinear dynamical systems using neural networks. In the proposed procedure, a neural model is firstly designed to simulate the desired nonlinear system. Then, based on the resulted simulation model, the new proposed observer is designed which insures the stability of the estimation error and at the same time it goes to zero faster than the time response of the system to be observed. The resulted estimated states can, therefore, be used to design a feedback controller for the system. Simulation examples are given and the obtained results reveal the effectiveness of the proposed technique.