Hybrid Approach for Modeling and Solving of Kinematics of a Compact Bionic Handling Assistant Manipulator

Othman Lakhal (1) , Achille Melingui (2) , Rochdi Merzouki (3, 4, 1)

1 CRIStAL - Centre de Recherche en Informatique, Signal et Automatique de Lille - UMR 9189
2 University of Yaoundé [Cameroun]
3 Polytech Lille - École polytechnique universitaire de Lille
4 MOCIS - Méthodes et Outils pour la Conception Intégrée de Systèmes

—This paper deals with a methodology for a real-time solving of a complex kinematics of a class of continuum manipu-lators, namely the Compact Bionic Handling Assistant (CBHA). First, a quantitative approach is used to model kinematically the CBHA inspired from the modeling of parallel rigid manipulators. For this case, the CBHA is modeled as a series of vertebrae, where each vertebra is connected to the next one through a flexible link. The latter named an inter-vertebra is modeled by a 3UPS-1UP (Universal-Prismatic-Spherical) joints. The kinematic models of the CBHA are derived from the Inverse Kinematic Equations (IKE) of each inter-vertebra. A qualitative approach based on neural networks is used to provide approximated solutions of the IKE for real-time implementation. Thus, the combination of the advantages of quantitative and qualitative approaches allows proposing a hybrid methodology for accurate modeling and solving the kinematics of this class of continuum robots. A set of experiments are conducted using a CBHA in order evaluate the level of efficiency of the proposed hybrid approach.

Domaines: Robotique [cs.RO]; Automatique; Modélisation et simulation; Systèmes et contrôle [cs.SY]

Othman Lakhal, Achille Melingui, Rochdi Merzouki. Hybrid Approach for Modeling and Solving of Kinematics of a Compact Bionic Handling Assistant Manipulator. IEEE/ASME Transactions on Mechatronics, 2016, 21 (3), pp.1326 - 1335. ⟨10.1109/TMECH.2015.2490180⟩. ⟨hal-01660738⟩