Repetitive predictive control and its application to PMSMs

Chai, S 2013, Repetitive predictive control and its application to PMSMs, Doctor of Philosophy (PhD), Electrical and Computer Engineering, RMIT University.

Document type: Thesis
Collection: Theses

Attached Files
Name Description MIMEType Size
Chai.pdf Thesis application/pdf 8.05MB
Title Repetitive predictive control and its application to PMSMs
Author(s) Chai, S
Year 2013
Abstract Repetitive Control is a learning control algorithm used to solve the problems of tracking the references and/or rejecting the disturbances that have repetitive nature. One of the challenging problems in repetitive control is to maintain the performance of the controller when the manipulated and/or state variables are hitting the constraints. Meanwhile, it is well known that Model Predictive Control (MPC) has its reputation in dealing with the constrained control problem through the use of optimization algorithms. This thesis incorporates the concept of repetitive control into the design of an MPC controller, resulting a new controller termed Repetitive-Predictive Control (RPC), so that the benefits of both controllers are combined, such as repetitiveness, constraints and multi-variable control. The design of the RPC controller is achieved by incorporating the dominant frequency components identified by the frequency decomposition of the reference signal into the receding horizon control of MPC. To further investigate the strength and weakness of the RPC, the design, tuning and performance of the RPC controller is thoroughly explored by its application to the control of Permanent Magnet Synchronous Motors (PMSMs) that have been broadly adopted for industrial control application due to their low volume and high efficiency.

The decision to use PMSMs as the application of RPC is reflected by the increasing trend to apply the Repetitive Control (RC) and Model Predictive Controller (MPC) for the electric drives in recent years. The design of RPC for the position, speed and current regulation of a PMSM has been investigated under two different schemes based on the Field Oriented Control (FOC). The first scheme employs the cascade structure with constrained MPC and RPC replacing the PI controllers for the inner-loop current control and outer-loop speed/position control, respectively. The second scheme is to combine both speed and current controllers into one single multi-variable model predictive controller with operating constraints imposed. The experimental comparisons of the two control schemes with cascade PI controllers demonstrate the superior performance of cascade RPC/MPC in terms of the ability of constrained control, disturbance rejection and position tracking. All results in the thesis have been validated by an experimental test-bed with an industrial-sized PMSM.
Degree Doctor of Philosophy (PhD)
Institution RMIT University
School, Department or Centre Electrical and Computer Engineering
Keyword(s) Repetitive Control
Model Predictive Control
Constrained Control
Resonant Control
Permanent Magnet Synchronous Motors
Electric Drives
Position Control
Disturbance Rejection
Version Filter Type
Access Statistics: 220 Abstract Views, 1362 File Downloads  -  Detailed Statistics
Created: Fri, 01 Nov 2013, 10:29:18 EST by Denise Paciocco
© 2014 RMIT Research Repository • Powered by Fez SoftwareContact us