Implementation of a power adaptive fuzzy control system for end milling processes.

Abstract

This thesis reports a fuzzy control system designed for power control of end milling processes. As compared to most of the existing end milling control systems, the proposed fuzzy control system has the following advantages: (a) multi-parameter adjustment; (b) insensitive to changes in work piece geometry, cutter immersion rate, and workpiece material; (c) cost-efficient and easy to implement; and (d) mathematically modeling-free. The proposed fuzzy controller is a two-input two-output system with simple triangular membership functions for both feedrate and spindle speed. The system also features a scaling factor adjustment mechanism used to tune the gain coefficients. The system is first examined by simulation using Simulink and Matlab fuzzy logic toolbox and then verified by various experiments on a CNC milling machine. The experiments carried out include: (i) steel cutting with different change patterns (gradual linear change, gradual curved change, and abrupt step change) in depth of cut; (ii) steel cutting with full and partial immersion rates; (iii) steel cutting with variable immersion rate; and (iv) aluminum cutting with step change in depth of cut. The experimental results show that as compared to single parameter (feedrate) adjustment the material removal rate can be improved by up to 25% when both feedrate and spindle speed are adjusted. It is also shown that the proposed system is stable and displays very good transient performance under all of our experimental conditions, indicating sound robustness. The use of power sensor has significantly reduced investment cost and avoided tedious setups. The simplicity in design and implementation of the system has been demonstrated in our development process.