Genetic Programming-Based Inverse Kinematics for Robotic Manipulators

Created by W.Langdon from gp-bibliography.bib Revision:1.7630

@InProceedings{Reuter:2022:EuroGP,

• author = "Julia Reuter and Christoph Steup and Sanaz Mostaghim",
• title = "Genetic Programming-Based Inverse Kinematics for Robotic Manipulators",
• booktitle = "EuroGP 2022: Proceedings of the 25th European Conference on Genetic Programming",
• year = "2022",
• editor = "Eric Medvet and Gisele Pappa and Bing Xue",
• series = "LNCS",
• volume = "13223",
• publisher = "Springer Verlag",
• address = "Madrid, Spain",
• pages = "130--145",
• month = "20-22 " # apr,
• organisation = "EvoStar, Species",
• keywords = "genetic algorithms, genetic programming, Cooperative Coevolution, Multi-Objective Optimization, Inverse Kinematics",
• isbn13 = "978-3-031-02055-1",
• DOI = "doi:10.1007/978-3-031-02056-8_9",
• abstract = "we introduce an inverse kinematics model for a robotic manipulator using Genetic Programming (GP). The underlying problem requires learning of multiple joint parameters of the manipulator to reach a desired position in the Cartesian space. We present a new approach to identify a closed-form solution for the Inverse Kinematics (IK) problem, namely IK-CCGP. The novelty of IK-CCGP is the cooperative coevolutionary learning strategy. Unlike other GP approaches, IK-CCGP is not limited to a certain angle combination to reach a given pose and is designed to achieve more flexibility in the learning process. Moreover, it can operate both as single- and multi-objective variants. In this paper, we investigate whether the inclusion of further objectives, i.e. correlation and the consistency of a solution with physical laws, contributes to the search process. Our experiments show that the combination of the two objectives, error and correlation, performs very well for the given problem and IK-CCGP performs the best on a kinematic unit of two joints. While our approach cannot attain the same accuracy as Artificial Neural Networks, it overcomes the explainability gap of IK models developed using ANNs.",
• notes = "http://www.evostar.org/2022/eurogp/ Part of \cite{Medvet:2022:GP} EuroGP'2022 held inconjunction with EvoApplications2022 EvoCOP2022 EvoMusArt2022",

}

Genetic Programming entries for Julia Reuter Christoph Steup Sanaz Mostaghim

Citations