Modeling and control design of a multifingered robot hand for object grasping and manipulation tasks

• Autor(a) principal: Reis, Matheus Ferreira dos
• Data de Publicação: 2016
• Tipo de documento: Trabalho de conclusão de curso
• Idioma: eng
• Título da fonte: Repositório Institucional da UFRJ
• Texto Completo: http://hdl.handle.net/11422/18198
• topic: CNPQ::ENGENHARIAS; Robot Modeling; Multifingered Robot Hands; Kinematic Control; Visual Servoing Control

In this work, we address the problem of kinematic modeling and control design of a multifingered robot hand. Each robot finger is modeled as a parallel manipulator and its kinematic constraints are computed from empirical analysis due to the inherent mechanical complexity of the mechanism. The motion control problem for a grasped object is solved by using the kinematic control approach, which is able to ensure the asymptotic stability of the output tracking error and appropriate prehension of the object. The kinematics-based control scheme is designed to include the contact model in the hand Jacobian matrix, allowing for the simultaneous control of the object position as well as the relative position between the fingers. Additionally, we also deal with the problem of controlling the 3D Cartesian position of an object of known geometry grasped by the robot hand using the visual servoing approach. A video camera and a depth sensor are simultaneously used to provide visual positioning and depth information of the grasped object. The design of the visual servoing controller is also based on the kinematic control approach, and two visual servo controllers are developed in the image space, ensuring the zero-convergence of the output tracking errors and high disturbance-rejection capability. Simulations and experiments are carried out with a three-fingered robot hand and a Microsoft Kinect sensor, illustrating the performance and effectiveness of the proposed methodology