Disturbance Rejection Using Disturbance Force Observer For XYZ Positioning Table

• Main Author: Jamaludin, Jailani
• Format: Thesis
• Language: English; English
• Published: 2015
• Subjects: T Technology (General); TJ Mechanical engineering and machinery
• Online Access: http://eprints.utem.edu.my/id/eprint/16882/1/Disturbance%20Rejection%20Using%20Disturbance%20Force%20Observer%20For%20XYZ%20Positioning%20Table.pdf; http://eprints.utem.edu.my/id/eprint/16882/2/Disturbance%20rejection%20using%20disturbance%20force%20observer%20for%20xyz%20positioning%20table.pdf; http://eprints.utem.edu.my/id/eprint/16882/; https://plh.utem.edu.my/cgi-bin/koha/opac-detail.pl?biblionumber=96104

In milling process, disturbance forces such as friction force and cutting force act directly on work surfaces thus producing external impact on the drives system of the positioning table. This effect must be compensated in order to preserve accuracy and quality of the final product. The characteristic of these cutting forces varied with cutting variables such as spindle speed, depth of cut, and feed rate. This thesis focuses on cutting force compensation using disturbance force observer (DFO) as an add-on control module to a classical and conventional PID control structure. DFO is a type of estimator that estimates precisely disturbance forces with prior knowledge of the cutting force harmonics frequencies content. The DFO was designed based on cutting forces measured from actual milling cutting process performed at constant depth of cut of 0.5mm, feed rate of 2000mm/min, and spindle speed of 1000rpm. These measured cutting data was applied as actual disturbance into the XYZ positioning milling table for numerical analysis and experimental validation of the estimator design. All numerical analysis was performed using MATLAB and Simulink software. In machine controller design, conventional controllers such as PID and cascade controller are widely used for positioning control. Therefore, based on this observation, four different control configurations were considered; PID controller, cascade P/PI controller, PID controller with inverse model based disturbance observer (IMBDO), and PID with DFO module. This thesis compares the performance of the PID controller that was combined with an observer, either IMBDO or DFO. Numerical and experimental analyses were performed using maximum tracking errors, root mean square (RMS) of the tracking errors, and magnitudes of the Fast Fourier Transform (FFT) of the tracking errors. Results obtained showed that PID with add-on DFO module produced superior performance against other control configurations. PID combined with DFO produced the most percentage reduction in maximum tracking error averaging at 93.76% for input disturbance frequencies of 5Hz, 15Hz and 35Hz. In comparison, cascade P/PI controller managed to record a 66.22% error reduction while PID with IMBDO produced a reduction of 52.75%. In term of RMS error, the experimental results showed that PID in combination with DFO produced the most percentage reduction that was 63.89% compared to the PID controller and PID with IMBDO and cascade P/PI where each produced a reduction of only 33.33%. Lastly, for FFT error analysis, the experimental results showed that PID combined with DFO produced the most improved performance with an average reduction of 74.87% at harmonic frequencies of 2.833Hz, 17.33Hz and 35.17Hz. This was in comparison to cascade P/PI controller and PID with IMBDO that produced a reduction of 44.39% and 21.97% respectively. Further studies on improvement of control performance especially in areas such as robustness and adaptivity of the control scheme to changing cutting conditions are desired.