Efficient Regulation of the Joint Velocities for the System Stability in Uncalibrated Visual Servoing Using Large Residual

Efficient Regulation of the Joint Velocities for the System Stability in Uncalibrated Visual Servoing Using Large Residual

  This paper addresses a target tracking system with estimation of the image Jacobian without the knowledge of camera configuration or robot kinematics. We propose an efficient target tracking algorithm based on a numerical technique. A method is also proposed to efficiently regulate the joint velocities so that the system stability is guaranteed. The robot system is controlled using the nonlinear least squares optimization technique. The full Newton’s method and the secant approximation method are used to calculate the joint angles. Using these algorithms joint values can be determined iteratively. Large residuals of joint values are determined using the secant approximation method. A secondary cost function is proposed with the modified iterative forms of the joint values in order to regulate the joint velocities. The image Jacobian is then estimated using the Broyden’s method and the recursive least squares (RLS) algorithm. The proposed method has been successfully applied in a visual servoing task.