Mahsa Haghani - Department of Electrical Eng., Faculty of Eng Islamic Azad University. Tehran
Abbas Ghanbari - Department of Electrical Eng., Faculty of Eng Islamic Azad University. Tehran
Mahmood Mola - Department of Electrical Eng., Faculty of Eng Islamic Azad University. Tehran

The main idea of chaos is defining behavior of certain systems that are highly sensitive to their initial condition. In chaotic systems, using a nonlinear controller is recommended to achieve good performance. Sliding mode controller is a robust state feedback controller to deal with uncertainties and has fast transient response. However, it is not an appropriate controller in the steady state of the system because of the existence of chattering. On the other hand, the neural controller is successfully applied on the control of chaotic systems but transient performance of this controller is not appropriate in the face of uncertainties due to its inevitable learning process In this paper, a hybrid sliding mode-neural network controller with variable weights is designed to utilize the advantages of both controllers. A fuzzy observer has the task of optimal tuning of these coefficients. The fuzzy observer also supervises the switching process between the two controllers.

formatting; style; styling; insert