Expert system based magnetic bearing control /
Active magnetic bearing (AMB) control systems have been around for many years. Contemporary controllers, such as PID controllers, are of the regulator format in which the desired rotor position is centered in the five axis bearing configuration. Since AMB systems have many nonlinear components suc...
| Main Author: | |
|---|---|
| Format: | Thesis Book |
| Language: | English |
| Published: |
[Place of publication not identified] :
[publisher not identified] ;
2001.
|
| Subjects: |
| Summary: | Active magnetic bearing (AMB) control systems have been around for many years. Contemporary controllers, such as PID controllers, are of the regulator format in which the desired rotor position is centered in the five axis bearing configuration. Since AMB systems have many nonlinear components such as the actuator force, material magnetic properties, and power amplifiers, conventional controllers are usually linearized about the centered position. Recently, some applications have been developed which require servo or non-zero reference inputs. Using a rotor as an actuator can have a variety of applications in turbomachinery such as active stall cell control and noise control. This thesis develops a servo controller based on direct linearization of the nonlinear actuator force. A state-space model was used to perform nonlinear simulations. In addition, a fuzzy logic based expert system is designed to improve performance. The expert system monitors criteria such as steady-state error, phase lag, and I²R power losses and modifies controller coefficients online. The controller and expert system were implemented with an AMB rotor kit developed by Revolve Magnetic Bearing, Inc., a host PC, and a stand alone digital signal processor (DSP). The five axis AMB rotor kit provided the real time inputs to the DSP and the expert system. The axial control algorithm resided on the DSP. The expert system was designed and implemented with Labview. The expert system, residing on the host computer, provides changes in controller coefficients, which are downloaded to the DSP operating in closed-loop mode. The theory and experimental results have shown that servo control of active magnetic bearings can be achieved while the fuzzy logic expert system was able to improve performance. |
|---|---|
| Item Description: | Vita. "Major subject: Mechanical Engineering". |
| Physical Description: | xi, 123 leaves : illustrations ; 28 cm. Issued also on microfiche from Lange Micrographics. |
| Bibliography: | Includes bibliographical references (leaves 111-112). |