Methods for Optimal Guidance of Industrial Robot Motions

Abstract

Industrial robots are used in a number of very different tasks. Typical applications are spot welding, arc welding, assembly, painting, material handling. Therefore application flexibility is a key concept for an industrial robot. Application flexibility of a robot is greatly enhanced by the use of fast and easy methods of teaching the robot. This paper deals with a new man-machine interface for teaching robots. At our institute a colour-screen-system with light pen has been developed. This system is used for the definition of the work flow at workpiece handling from a moving conveyor to the place of deposit. With this system the operator can draw instructions over a real video frame showing the robot and the working place. Possible instructions are obstacles or intermediate positions of the robot path. A first realization was done with one camera at the ceiling.A further crucial point treated in this paper is the calculation of robot paths. Depending on intermediate path points, forbidden areas and given velocities the paths of robot motions are computed. The computation algorithm gets its input data from the teaching informations as well as from an optical sensor, which recognizes position and orientation of a workpiece on a moving conveyor. Because the calculation of robot paths is done in joint coordinates transformations from “world ”-coordinate systems (teaching and sensor coordinate systems) into the joint coordinate system are necessary. For the trajectories of the robot motions polynomials are determined. Determining these polynomials various optimisation criteria (length, time, energy consumption, minimum stress imposed on material) can be taken into consideration. The calculations for the optimal guidance of robots are done by a microprocessor, which is a part of a distributed microprocessor system.