Research of real time robot visual servoing based on genetic algorithm

This paper presents a position-based approach to visual servoing and pose estimation for an unknown target where a robot manipulator works in look-and-move mode to reach the object. This approach employs the step genetic algorithm and a pattern matching technique to search for a target object in the robot task space. A GA-based controller is presented and a stability proof is given. Simulation results performed on a two-link robot manipulator show that the manipulator hand can reach the object and search for an unknown target quickly. It is proved that this approach can be used in real-time control of a manipulator to pick up an object whose position and orientation are not known previously.