Combined learning and identification techniques in the control of robotic manipulators

摘要

A nonlinear analysis of iterative learning control procedures is carried out for robotic systems, without assuming the presence of any online acceleration feedback. Global sufficient conditions for stability and convergence of the procedure are presented in terms of the accuracy with which the inertia matrix of the robot must be known. An algorithm for the identification of the parameters of the matrix, capable of guaranteeing the required accuracy, is presented.< <ETX xmlns:mml="http://www.w3.org/1998/Math/MathML" xmlns:xlink="http://www.w3.org/1999/xlink">&gt;</ETX>