Asymptotically stable visual servoing of manipulators via neural networks

In this article we present a class of position control schemes for robot ma nipulators based on feedback of visual information processed through artifi cial neural networks. We exploit the approximation capabilities of neural n etworks to avoid the computation of the robot inverse kinematics as well as the inverse task space-camera mapping which involves tedious calibration p rocedures. Our main stability result establishes rigorously that in spite o f the neural network giving an approximation of these mappings, the closed- loop system including the robot nonlinear dynamics is locally asymptoticall y stable provided that the Jacobian of the neural network is nonsingular. T he feasibility of the proposed neural controller is illustrated through exp eriments on a planar robot. (C) 2000 John Wiley & Sons, Inc.