Visual servoing of a robotic manipulator using an optimized trajectory planning technique

摘要

In this paper, a novel image based visual servoing (IBVS) using an optimized trajectory planning technique is proposed. In this method the camera's velocity screw is parameterized using polynomial profiles. The parameters of the velocity profile are then determined to guide the robot to its desired position, by minimizing the error in an optimization algorithm. In addition, this technique keeps the robot in its physical limits and keeps the features in the field of view (FOV). This algorithm is tested on a 4 DOFs Denso robot in an eye-in-hand configuration. Simulation results demonstrate t hat the proposed method solves some of the IBVS problems such as reaching the robot's or cameras' limits and instability of the system for pure 180o rotation about its center.