الفهرس 
Support Vector MachineOnly 14 pages are availabe for public view
 |  | from | 255 |  |  |
| | | from | 255 | | |
Abstract
In this thesis the performance of the intelligent controllers is improved using the machine learning approach to control the nonlinear systems where the nonlinear systems have inherent uncertainties in terms of precision of the sensors, a noise produced by the sensors, nonlinear characteristic of the actuator and the system structure. In this thesis, the reinforcement learning is developed from the machine learning approach to get stable controllers for nonlinear systems. The main advantages of using reinforcement actor-critic learning for controlling the systems are that it acquires experience of the system during the operation and it does not need the system model. So it gives high performance for nonlinear and time varying parameters of the system