The configuration space transformation for articulated manipulators: a novel approach based on RBF-networks

This paper proposes a neural network-based method for the computation of the configuration space for robotic manipulators. The configuration space can be obtained by repeatedly computing configuration space patterns for elementary obstacle primitives. For any manipulator, these patterns depend only on the distance between the base of the manipulator and the obstacle primitive. An RBF-network is trained to recognise the distance of an obstacle primitive and to respond with the associated pattern. The interpolating features of radial basis functions are exploited to achieve a good approximation for untrained patterns. The main principles of the method are explained for a two-link manipulator. Training results are reported.