A state observer-based tracking controller for suppression of input disturbance in machine tools application
In milling process, disturbance forces such as cutting force and friction force act directly on the servo drive system producing unwarranted effect that deteriorates the accuracy of the positioning table. This effect has to be compensated in order to preserve geometrical accuracy and quality of the...
Saved in:
| Main Author: | |
|---|---|
| Format: | Thesis |
| Language: | English English |
| Published: |
2021
|
| Subjects: | |
| Online Access: | http://eprints.utem.edu.my/id/eprint/26079/1/A%20state%20observer-based%20tracking%20controller%20for%20suppression%20of%20input%20disturbance%20in%20machine%20tools%20application.pdf http://eprints.utem.edu.my/id/eprint/26079/2/A%20state%20observer-based%20tracking%20controller%20for%20suppression%20of%20input%20disturbance%20in%20machine%20tools%20application.pdf |
| Tags: |
Add Tag
No Tags, Be the first to tag this record!
|
| Summary: | In milling process, disturbance forces such as cutting force and friction force act directly on the servo drive system producing unwarranted effect that deteriorates the accuracy of the positioning table. This effect has to be compensated in order to preserve geometrical accuracy and quality of the final product. This thesis focuses on suppression of disturbance force characterise by harmonic frequencies dictating by the spindle speed of the milling table using state observer-based controller for precise tracking performances of the motion drive system. This thesis proposes improvement to control performance of classical cascade P/PI controller via add-on modules to the control structure consisting of state observers named inverse model-based disturbance observer (IMBDO) and disturbance force observer (DFO). The cascade P/PI controller was designed using traditional loop shaping frequency domain method. IMBDO estimates the input disturbance and any unmodeled system dynamics while DFO performs direct estimation of the cutting force using information of harmonic frequencies corresponding to the sinusoidal based input disturbance force. Numerical analysis was performed using MATLAB/Simulink software and experimental analysis was performed on the x-axis of an XY milling positioning table ball screw driven system. This thesis compares the performance of cascade P/PI with add-on IMBDO plus DFO with other control configurations; (i) a cascade P/PI stand-alone, (ii) cascade P/PI with IMBDO, and (iii) cascade P/PI with DFO. The control performances of these configurations were analysed using maximum tracking errors (MTE), root mean square (RMSE) of the tracking errors, and magnitudes of the Fast Fourier Transform (FFT) of the tracking errors. Results obtained showed that cascade P/PI with add-on IMBDO plus DFO module produced superior performance against other control configurations. Maximum tracking error results showed that cascade P/PI with IMBDO plus DFO produced the best tracking performances for all harmonic frequencies considered yielding percentage errors reduction of 97.52%, 98.70% and 99.13% for input disturbance of one harmonic, two harmonics, and three harmonics respectively. In term of RMSE values, the experimental results showed that cascade P/PI with IMBDO plus DFO produced the most percentage error reduction with values recorded at 98.80%, 97.75% and 97.97% for the respective input harmonics. In term of FFT results, cascade P/PI with IMBDO plus DFO produced the most reduction in peak amplitudes with values corresponding to 99.78% for the first harmonic, 99.67% and 99.53% for the second harmonics and 99.86%, 99.81% and 99.91% for the third harmonics. The closed loop and sensitivity transfer function of this control configuration confirmed the superiority of cascade P/PI with IMBDO plus DFO in yielding the smallest tracking error thus yielding the most efficient positioning control system. |
|---|