Amélioration de la précision des robots industriels pour des applications d'usinage a grande vitesse

Abstract

Industrial robots are usually used to realize industrial tasks like material handling, welding, cutting and spray painting. The mobility, flexibility and important work space of these robots allow using them in new fields of industry such as prototyping, cleaning and pre-machining of casts parts as well as end-machining of middle tolerance parts. These new applications require a high level pose accuracy and to achieve a good path tracking. Unfortunately industrial robots were designed to have a good repeatability but not a good accuracy. Their poor accuracy is caused by geometric factors and non-geometric factors. Many fields of investigation are proposed to increase the accuracy of industrial robots like: robot calibration, process development and control system. In this work, firstly, the focus lies on developing an efficient and pragmatic trajectory planning algorithm for continuous machining operations with a 6-axis robot. Secondly, compliances of robots joints are identified and used to calculate the Cartesian compliance at the end effector of the robot. This Cartesian compliance is employed in a strategy to optimize the stiffness of robot structure during machining. The compliances of robot joints are used also in an elasto-static model allowing anticipating the displacement of the end-effector caused by external loads. Finally, a model based technique is used to compensate two joints major errors: compliance error and kinematic errors. The model of each error is integrated then in our trajectory planner in order to realize an off-line compensation of joints errors.