Adaptive Controller Design for Cutting Forces Rejection in Machine Tools

Lim, Hong Xuan (2024) Adaptive Controller Design for Cutting Forces Rejection in Machine Tools. Final Year Project (Bachelor), Tunku Abdul Rahman University of Management and Technology.

Text LIM HONG XUAN_Full Text.pdf Restricted to Registered users only Download (4MB)

Abstract

In the domain of manufacturing, particularly within the sector of milling operations, precision and efficiency are paramount. Electromechanical actuators (EMAs) are increasingly employed in modern milling machines due to their ability to provide high-speed and precise positioning capabilities. This thesis addresses the challenge of poor tracking performance caused by dynamic cutting forces during the milling process. An in-depth investigation into the effects of varying spindle speeds (1000, 1500, and 2000 rpm) on the disturbance characteristics introduced by cutting forces is presented. To enhance the disturbance rejection and tracking accuracy of the milling system, this study develops and evaluates three advanced control strategies: a Proportional- Integral-Derivative (PID) controller, a Sliding Mode Control (SMC) controller, and an Active Disturbance Rejection Control (ADRC) strategy. Each controller was integrated into a singleaxis EMA-based milling setup and assessed through comprehensive simulation models. The effectiveness of these controllers was quantitatively analyzed through metrics such as Maximum Tracking Error (MTE) and Root Mean Square Error (RMSE), across the stated spindle speeds. Notably, ADRC achieved a reduction in MTE by up to 71.56% at 1000 rpm and maintained high-performance levels at higher spindle speeds, with improvements of 71.27% and 71.57% at 1500 rpm and 2000 rpm respectively. The ADRC is able to maintain a consistent RMSE of 0.0178 mm, indicating a 70.82% improvement, showcasing its robustness against disturbances introduced by the cutting force.The ADRC demonstrated superior performance in both disturbance rejection and tracking error reduction across all tested conditions. Notably, ADRC consistently showed significant improvements of 95.86% and 98.59% in harmonics reduction at lower frequencies such as 16 Hz and 32 Hz, a critical factor for enhancing the quality of the milling process. Future work is suggested to include real-world implementation of these controllers to validate simulation results, extend the analysis to multiple axes, and explore the integration of machine learning techniques for predictive disturbance compensation. This research contributes to the body of knowledge in adaptive control strategies for milling operations, proposing robust solutions that enhance operational accuracy and efficiency in industrial manufacturing environments.