Representation and learning of nonlinear compliance using neural nets

Abstract

A new approach to compliant motion control using neural networks is presented. In the paper, "compliance" is treated as a nonlinear mapping from a measured force to a corrected motion. The nonlinear mapping by a multilayer neural network is outlined, this allows one to deal with complex control strategies that cannot be represented by linear compliance, such as in stiffness and damping control.< <ETX xmlns:mml="http://www.w3.org/1998/Math/MathML" xmlns:xlink="http://www.w3.org/1999/xlink">&gt;</ETX>