الفهرس 
IntroductionOnly 14 pages are availabe for public view
 |  | from | 97 |  |  |
| | | from | 97 | | |
Abstract
In the presence of nonlinear system properties and unknown parameters, designing a controller to meet certain performance requirements became a very difficult task. Controlling these uncertain nonlinear systems is a challenging task in the field of control systems. This thesis presents a hybrid controller, which is a backstepping controller based on an adaptive Elman neural network (BSAENN) to solve the uncertainty problem of underactuated robotic systems. This controller is adapted using a combination of the adaptive Elman neural network (AENN), which is a recurrent multilayered neural network, and the traditional backstepping control (TBS) approach. For this reason, we didn’t consider this controller. The AENN with the adaptive parameters is exploited to approximate the unknown nonlinear system uncertainties and enhance the control behavior against uncertainties. Approximation of the unknown nonlinear system dynamics and uncertainties can be achieved using methods such as wavelets neural networks, or combination between neural networks and fuzzy control. Using fuzzy controller as an approximator requires a large number of fuzzy rules, which leads to a sluggish system. The proposed controller has been applied to a nonlinear underactuated robotic system. The stability of the closed-loop system is proved using Lyapunov stability theory. Numerical simulations with dynamical model of the two link robot, compared to conventional controllers (PID and TBS), show that the proposed controller provides robustness for trajectory tracking performance under the occurrence of uncertainties.