Abstract

The ability to provide a physical interaction between an industrial robot and a workpiece in the environment is essential for a successful manipulation task. In this context, a wide range of operations such as deburring, pushing, and polishing are considered. The key factor to successfully accomplish such operations by a robot is to simultaneously control the position of the tool-tip of the end-effector and interaction force between the tool and the workpiece, which is a challenging task. This thesis aims to develop new reliable control strategies combining vision and force feedbacks to track a path on the workpiece while controlling the contacting force. In order to fulfill this task, the novel robust hybrid vision and force control approaches are presented for industrial robots subject to uncertainties and interacting with unknown workpieces.
The main contributions of this thesis lie in several parts. In the first part of the thesis, a robust cascade vision and force approach is suggested to control industrial robots interacting with unknown workpieces considering model uncertainties. This cascade structure, consisting of an inner vision loop and an outer force loop, avoids the conflict between the force and vision control in traditional hybrid methods without decoupling force and vision systems. In the second part of the thesis, a novel image-based task-sequence/path planning scheme coupled with a robust vision and force control method for solving the multi-task operation problem of an eye-in-hand (EIH) industrial robot interacting with a workpiece is suggested. Each task is defined as tracking a predefined path or positioning to a single point on the workpiece’s surface with a desired interacting force signal, i.e., interaction with the workpiece. The proposed method suggests an optimal task sequence planning scheme to carry out all the tasks and an optimal path planning method to generate a collision-free path between the tasks, i.e., when the robot performs free-motion (pure vision control). In the third part of the project, a novel multi-stage method for robust hybrid vision and force control of industrial robots, subject to model uncertainties is proposed. It aims to improve the performance of the three phases of the control process: a) free-motion using the image-based visual servoing (IBVS) before the interaction with the workpiece; b) the moment that the end-effector touches the workpiece; and c) hybrid vision and force control during the interaction with the workpiece. In the fourth part of the thesis, a novel approach for hybrid vision and force control of eye-in-hand industrial robots is presented which addresses the problem of camera’s field-of-view (FOV) limitation. The merit of the proposed method is that it is capable of expanding the workpiece for eye-in-hand industrial robots to cope with the FOV limitation of the interaction tasks on the workpiece. All the developed algorithms in the thesis are validated via tests on a 6-DOF Denso robot in an eye-in-hand configuration.