Tracking with a fuzzy logic controller as applied to a pneumatic robot for polishing

Abstract

This paper examines the performance of a fuzzy logic controller (FLC) as implemented on a pneumatic robot for application to the polishing of sheet steel. A 3-axis gantry robot with rodless pneumatic cylinders for actuation and proportional flow control valves was modified to enable experiments on the polishing of steel sheet. An air motor with a buffing wheel was mounted on the Z-axis to provide the polishing action, with the Z-axis cylinder controlling the applied buffing force. Based upon past experience, the FLC for X and Y-axis trajectory control. was supplemented by both system time lag and friction compensators with velocity feedback. In this application, performance is measured by the controller's ability to track a two dimensional shape as well as its ability to minimize the absolute tracking error. Details of the FLC controller including the nature of the membership functions and the rule base are given. Experimental results show that: a) FLC tracks better than a traditional proportional-velocity-differential pressure (PVdP) controller, b) lag and friction compensators are effective and c) the zig-zag trajectory needed for polishing is possible. Future work includes confirming that FLC is more robust than PVdP, in that it should be able to maintain tracking performance as the polishing wheel is applied to the sheet steel, which will generate a variable load.