الفهرس 
AbstractOnly 14 pages are availabe for public view
 |  | from | 153 |  |  |
| | | from | 153 | | |
Abstract
The Induction Motor (IM) is the horse power of industry. It covers a large section of the motor used because of its advantages compared to other motors like DC- and synchronous motors. It has a simple construction design, ruggedness in operation, and low maintenance. Despite that, it has a complex non-linear model, rotor variables are rarely measurable, its parameters vary with operating conditions, and the coupling between the different variables of control, which make a challenge the control of its speed. The Field Oriented Control (FOC) techniques have a great attention in the last decades. Since it solves the problem of interaction between motor variables and make the speed control of induction motor is similar to speed control of separately excited DC- motors. But these techniques have their own problems, such as, motor’s parameter mismatch between the modeled and actual values. One of these techniques is Indirect Rotor Field Oriented Control (IRFOC), which can be accomplished based on various ways. The other issue for IM speed control is that the speed and the flux are needed to be feedback to the controller. This can be achieved using sensors but it makes more costs on the control system and may be not suitable for some applications or get them by estimation. Nowadays there are many researches in speed and flux estimation methods each one has advantages and problems, this makes the field of search is still opened to solve these problems. This thesis presents two different techniques for IRFOC of IM based on Proportional Integral (PI) and Sliding Mode Control (SMC). The PI controllers are designed and there gains are tuned by two ways; conventional way based on “try and error”, and a proposed technique based on Particle Swarm Optimization (PSO) algorithm. The results are taken for the two techniques and compared. The SM controller is proposed to overcome the sensitivity of the control system to the parameters mismatch, the results show the superiority of the SMC to PI control in speed control of IM and its robustness against the variation of the system dynamics. A speed estimation algorithm is proposed based on the Artificial Neural Network (ANN), because of its advantage to deal with the nonlinear system. The model is approached with the PI and SM controllers. The results show a good response and matching of the estimated speed with the actual one for the two controllers especially for the SMC which has excellent results compared to the PI controller at low speed operation.