Iterative Learning Control for A Class of Robot Manipulators Tang Jianzhong, Zhang Yangjun
(Mechanical Engineering School, Xi'an Jiaotong University, Xi'an, China£©

Abstract: A new method of iterative learning control is presented for a class of robot mani- pulators. In contrast to other known methods of iterative learning control, the proposed method can utilise dynamic parameters of the drive device to determine the learning gain terms if the inertia of drive device translated to joint is enough large. For majority industrial robot manipulators, this condition can be satisfied because they usually have higher transmission ration between drive device and joint. Therefore, it is avoided to determine the learning gain terms through objectivity. An experiment for a robot mainpulator with three degree free is carried out to show effectiveness of the method, and via experiment, it is demonstrated that the method yeilds high learning speed and high control accuracy.
Keywords: robot manipulator, learning control, nonlinear .