Abstract
In this paper, we propose a system for sensing external world using a hand-eye camera for future autonomous guidance control on street. A hand-eye camera attached on a hand of a manipulator is useful comparing with a position-fixed camera, because it can obtain information of every desired direction by the motion of the manipulator. However, it will be difficult to fix the view axis of the camera to objects during traveling, because of existence or irregularities of street. A control of motion of the hand-eye camera on a car can be treated as the control system of mobile operations by the mobile manipulator. First, a model for a mobile manipulator traveling on irregular roads is formulated. Second, a neuro adaptive control system is proposed. Third, simulations are performed. Finally, we show the proposed system can decrease control errors caused by irregular roads.