Adaptive control of deburring robots based on human demonstration data

An adaptive controller for deburring robots is presented. The adaptation mechanism that updates the controller parameters is designed based on a set of teaching data taken from human demonstrations. The relationship between what a human monitors in a deburring process and how the human responds to variations in the task process is described as an associative mapping. This associative mapping can be recovered from human teaching data. The consistency of the mapping relation is analyzed by using Lipschitz´s condition. The associative mapping is used in this new adaptive controller as the scheme to compute and update control parameters based on burr characteristics. To take into account the surface finish of the workpiece as an explicit control objective, an integral feedback scheme is used. The control system is implemented on a direct drive robot. Experimental results show that the robot can adapt to the deburring process in a manner very similar to a human worker. With the use of an integral feedback action, the smoothness of the surface finish can be greatly improved