A Robust Trajectory Tracking Control of a Polishing Robot System Based on Cam Data

Polishing a die that has free-form surfaces is a time-consuming and tedious job, and requires a considerable amount of high-precision skill. In order to reduce the polishing time and cope with the shortage of skilled workers, a user-friendly automatic polishing system was developed in this research. The polishing system is composed of two subsystems, a three-axis machining center and a two-axis polishing robot. The developed polishing system with five degrees of freedom is able to keep the polishing tool normal to the die surface. To solve the problem of tracking errors related to the unmodeled dynamics in the operation of the polishing robot, a sliding mode control algorithm with velocity compensation, which is known to be robust against parameter variations and payload changes, is proposed. Trajectory tracking experiments showed that the effects of reducing the tracking error by the proposed sliding mode control is superior to that by the PD (proportional derivative) control. The polishing data is generated from CAD (computer aided design) data or from teaching data by PolyCAM, a CAM (computer aided manufacturing) system dedicated to the polishing system. This CAM system consists of the following four modules: a geometric modeller, a CAD data exchange module, a polishing data generation module, and a graphic simulator. In order to evaluate the performance of the polishing robot system, some polishing experiments on a shadow-mask die were performed. INTRODUCTION In the process of die manufacturing, some polishing must be performed to remove tool marks and to improve the smoothness and flatness of die surfaces. However, while the cutting process has been automated by the progress of CNC (computer numerical control) and CAD/CAM, the polishing process still depends on the experienced knowledge of an expert. Also, even when workers are skilled in polishing dies, it takes much time to obtain the required roughness and smoothness on the surface of a die. Moreover, many workers gradually avoid doing polishing work because of the poor working conditions caused by dust and noise. Therefore, to improve productivity and to solve the potential shortage of skilled workers, several studies on the automation of the polishing process have been conducted. [5,11-13] In this study a polishing robot with two degrees of freedom was developed. The robot has a pneumatic system and is attached to a machining center with three degrees of freedom, which is commonly used in industry. Therefore, the polishing system has a total of five degrees of freedom, and the system is able to keep the polishing tool normal to the die surface and can maintain a constant pressure. The PID (proportional integral derivative) control algorithm has been the most common type of industrial robot controller. However, the algorithm cannot provide high precision in high-speed tasks where sudden changes in parameters are frequent. Unless nonlinearities of robotic manipulators are compensated for properly, satisfactory control performance cannot be expected from the PID algorithm. Moreover, accurate modeling of robotic systems is very difficult, due to the nonlinear friction effect and changes in the payload during task executions. [3] To solve tracking errors related to the unmodeled dynamics in the operation of a polishing robot, this study proposes a sliding mode control algorithm that compensates for velocity. Also, to investigate the trajectory tracking performance of the proposed sliding mode control algorithm, the algorithm is compared with the PD control algorithm, which has been the most common type of industrial robot controllers. PolyCAM, a dedicated CAM software for the system, was developed to generate the desired polishing data. In order to evaluate the performance of the polishing robot system, some polishing experiments on a shadow-mask die were conducted. DESIGN OF SLIDING MODE CONTROL Sliding Mode Control The structure and the appearance of the two-axis polishing robot developed in this study are shown in Figure 1. The dynamic equation of the polishing robot can be written as follows: (2.1) where Ji is the summation of all linear terms in the moment of inertia of link i and the driving motor. Bi is the equivalent damping coefficient from the motor, reduction gears, and the viscosity friction of link i. The disturbance term Fi is the summation of the nonlinear terms: inertia moments, the Coriolis and centrifugal forces, the gravity force, and the Coulomb friction term. ki is a constant coefficient to