With the rapid development of automation technology, mechanical arms are widely used in various automated operations such as handling, transportation, welding, spraying, assembling and so on. To further improve the work efficiency of mechanical arms, this dissertation deeply studies the key techniques in the automatic programming of industrial robots both with and without three dimensional model of the object to be processed. The main work and innovation points can be summarized as follows: 1. Aiming at automatic programming applications of industrial robots based on three dimensional model of the object, a path planning technique based on region division is proposed. This technique can plan better paths in each region and generate safe transitional points between regions at the same time. As a result, the difficulty in model analysis and the complexity in path planning are greatly reduced. 2. For automatic programming applications of industrial robots based on three dimensional model of the object, a method in which the planned paths are converted to the final real control instructions, as well as an instruction modification technique based on simulation is put forward. Meanwhile, we apply the proposed method to the automatic spraying process based on three dimensional model. Simulations and experiments have demonstrated the validity of the proposed method. 3. For automatic programming applications of industrial robots based on three dimensional model of the object, an instruction reusing method is proposed. Based on human-computer interaction and improved ICP algorithm, the proposed method can get the pose deviation efficiently between the processing model and the same model in the library. On this basis, the work instructions can be got rapidly through converting of the instructions in the library. Simulations and experiments verify the validity of the presented technique. 4. Robot automatic programming technique without a 3D model of the object is studied. For automatic programming of a lower limb rehabilitation robot, a gait data acquisition method based on binocular stereo vision, as well as a gait curve generation technique is proposed, on the foundation of which automatic programming of lower limb rehabilitation robot is realized. Simulations and real experiments verify the validity of the presented method. 5. For automatic programming of lower limb rehabilitation robot, a virtual system with both off-line and on-line function is develop...
修改评论