engleski

Comparison of uncalibrated model-free visual servoing methods for small amplitude movement : a simulation study

The paper compares performance of several methods used for estimation of image Jacobian matrix in uncalibrated model-free visual servoing for eye- in-hand configuration with small movement amplitudes for several sets of system parameters. These methods include Broyden algorithm, Kalman and Particle filters as well as recently proposed population-based algorithm. Algorithms were tested in simulation environment (Peter Corke’s Robotic Toolbox for Matlab) on PUMA 560 robot for several application scenarios including static point and dynamic trajectory tracking for several characteristic shapes and three different speeds. Based on obtained results conclusions were drawn about strengths and weaknesses of each method in particular setup and in general, as well as on how using algorithm switching might improve overall robot tracking performance with respect to desired trajectory. Finally, possible future research directions are suggested.

visual servoing; model-free; uncalibrated; population-based; particle filter

nije evidentirano