Research On The Accuracy Assessment Anderror Compensation Of Industrial Robots

As modern industrial production equipments, industrial robots make use of multi-disciplinary knowledge. Growing number of manufactures use industrial robots instead of manual labor because of its high efficiency and reliability. Nowadays, industrial robots’ repeat positioning accuracy is high while its absolute positioning accuracy is low. Accuracy evaluation and error compensation of industrial robots, which is of great significance in the life of the robot and the expansion of the scope of robot application, can improve its absolute position accuracy.Firstly we study and summarize the domestic and foreign theory of industrial robot accuracy evaluation and error compensation. And the purpose and content of the articles is presented. Then we demonstrate the accuracy evaluation criteria and testing methods according to the national standards, ends pose error model of the robot is established by a modified five-parameter model, the identification of the geometric parameters is completed by using singular value decomposition combined with the least square iterative, next we use the Newton-Raphson iterative method to acquire compensation angle at all calibration configurations based on the error compensation hypothesis. Meanwhile, in order to solve the problem of the coefficient matrix perturbation, we solve the optimization model of matrix condition number by using genetic algorithm and get the optimal configurations of robot. Thirdly, we analyze the effect of all D-H parameters error on the accuracy of each link, end position and error models, and conclude that error model can be used to analyze the robots micro geometry error. And the simulation of geometric parameter error identification and error compensation is made to validate the proposed method. A software for robot accuracy assessment and error compensation is developed on VC6.0 platform to complete the accuracy evaluation and error compensation fastly and conveniently. Lastly, we complete the experiment for laboratory IRB120 by the entire robot position measurement system and calibration software in 35 minutes, which improves the absolute positioning accuracy from 0.5mm before calibration up to 0.3mm after calibration.The software we developed can be compatible with ABB robot controller. The real-time control of robots for calibration can be completed on the PC by serial communication of data. This research has some reference value for the software’s development of motion control and calibration.