Nonlinear optimization methods for uncalibrated robot visual servoing

-This paper considers the uncalibrated, model free, robot visual servoing as a nonlinear optimization problem in which stochasticity is present due to a noise and there is no analytical form of the objective function. Such (robot) systems are suitable for unknown and unstructure envinronments due to minimal requirements related to calibration and robot kinematic’s parameters. The numerical quasy-Newton methods offer a theoretical background for problem solving. However, additional attention has to be paid which assured stability and the robustness of the proposed method. In this paper we present the simulation results of various, well-known iterative numerical solution efficacy performing the dynamic target visual servoing. The results show that the generalized form of solutions perform better than the simple form adopted for a specific type of a goal functions. Key-Words: visual servoing, uncalibrated system, robot control, nonlinear optimization, generalized solutions