الفهرس 
Types of Adaptive ControlOnly 14 pages are availabe for public view
 |  | from | 194 |  |  |
| | | from | 194 | | |
Abstract
Many important processes in chemical engineering are inherently nonlinear. In spite of these nonlinearities, most industrial processes are regulated with simple ProportionalIntegralDerivative (PID) controllers designed using linear system analysis. Clearly, the use of linear control techniques is quite limited if a chemical process is highly nonlinear. The use of nonlinear process models within the control strategy has been shown to provide the potential for significant improvement over linear controllers for nonlinear processes. However, the mathematical complexity and large computational effort needed by nonlinear predictive control approach have limited its practical applications. In previous work, identification and control of nonlinear processes with PID controller has resulted in poor performance .To overcome these problems, this research concerns the development of a control strategy that is within the framework of control using neural networks. The ability of artificial neural networks to represent nonlinear systems make them a powerful tool for process modelling and control, the ability of Kohonen neural networks to predict the process response multiple steps ahead is assessed. The neural network model is then employed in a predictive control scheme which uses a kohonen algorithm. After this, the Kohonen network is modified by controlling number of units in hidden layer, and then applied on the CS.