Permanent magnet servo control systems using artificial neural network
This thesis investigates the enhancement of the permanent magnet servo systems through artificial neural network (ANNs). For accurate speed control, a P+RBFESO controller which combines the Radial Basis Function Neural Network (RBFNN) with the ESO-based ADRC is proposed to enhance the disturbance re...
| Main Author: | |
|---|---|
| Other Authors: | |
| Format: | Thesis-Master by Coursework |
| Language: | English |
| Published: |
Nanyang Technological University
2023
|
| Subjects: | |
| Online Access: | https://hdl.handle.net/10356/166803 |
| Summary: | This thesis investigates the enhancement of the permanent magnet servo systems through artificial neural network (ANNs). For accurate speed control, a P+RBFESO controller which combines the Radial Basis Function Neural Network (RBFNN) with the ESO-based ADRC is proposed to enhance the disturbance rejection capability.
Instead of fixed weights and biases, online learning is adopted to allow the control
system to maintain optimal performance in different operating conditions. This study provides a systematic presentation of the development and implementation of the proposed P+RBFESO controller. The effectiveness of the proposed solution is evaluated thoroughly through experiments, and the performance is compared with the conventional control methods. The results prove that the proposed P+RBFESO controller offers enhanced robustness and flexibility then their conventional counterparts. |
|---|