International Conference on Artificial Neural Networks (ICANN), 2002, Madrid (España) ; The main drawback of using neural networks to approximate the inverse kinematics (IK) of robot arms is the high number of training samples (i.e., robot movements) required to attain an acceptable precision. We propose here a trick, valid for most industrial robots, that greatly reduces the number of movements needed to learn or relearn the IK to a given accuracy. This trick consists in expressing the IK as a composition of learnable functions, each having half the dimensionality of the original mapping. A training scheme to learn these component functions is also proposed. Experimental results obtained by using PSOMs, with and without the decomposition, show that the time savings granted by the proposed scheme grow polynomically with the precision required. ; This work was supported by the project 'Computación mediante restricciones en robótica y gestión de recursos' (070-725). ; Peer Reviewed


    Access

    Download


    Export, share and cite



    Title :

    Learning inverse kinematics via cross-point function decomposition


    Contributors:

    Publication date :

    2002-01-01



    Type of media :

    Conference paper


    Type of material :

    Electronic Resource


    Language :

    English



    Classification :

    DDC:    629



    Learning inverse kinematics: Reduced sampling through decomposition into virtual robots

    Ruiz de Angulo, Vicente / Torras, Carme | BASE | 2008

    Free access

    Inverse Kinematics

    Frigeni, Fabrizio | Springer Verlag | 2022


    Inverse Kinematics

    Jazar, Reza N. | Springer Verlag | 2021


    Prioritized Inverse Kinematics: Generalization

    An, Sang-ik / Lee, Dongheui | German Aerospace Center (DLR) | 2019

    Free access

    An Inverse Kinematics Solution for Flexible Manipulators Using Learning Algorithm

    Uchiyama, M. / Usui, K. / Dai, Y. | British Library Online Contents | 1997