Vision-based hybrid control strategy for autonomous docking of a mobile robot

Parking or docking behavior is a very important behavioral unit for mobile robots. In a typical mobile robot application, robots have to navigate to a pre-defined destination and perform a given task. In order to perform those types of tasks, robots usually need to be precisely aligned against a target pose. In a vision-based mobile robot docking system, the nonholonomic nature and limited field-of-view constraints are required to be incorporated in the control strategy. In this paper we propose a novel hybrid control solution to this problem using image-plane measurements. The final docking position of the robot is defined as an image-plane reference. The robot is controlled to match the current view of the docking station to the reference view. The experimental results are used to demonstrate the effectiveness of the proposed approach. The results show that the proposed method provides a high level of robustness and precision which, is required in most practical applications.