Parameter estimation and object gripping based on fingertip force/torque sensors. (July 2021)

• Record Type: Journal Article
• Authors: Wang, Chao; Zang, Xizhe; Zhang, Xuehe; Liu, Yubin; Zhao, Jie
• Is Part Of: Measurement. Volume 179(2021)
• Publication Date: 2021-07
• Subjects: Robot system -- Fingertip force/torque sensor -- Task object -- Parameter estimation -- Gripping control; Weights and measures -- Periodicals; Measurement -- Periodicals; Measurement; Weights and measures; Periodicals; 530.8
• DOI: 10.1016/j.measurement.2021.109479
• Languages: English
• ISSNs: 0263-2241

Abstract

Highlights: The robot estimates the parameters of varying objects and grips them stably. The object's mass and barycenter are estimated via fingertip force/torque sensors. The effects of multi-dimensional force coupling and gripping force are eliminated. Adaptive adjustment of gripping force based on the object's parameters. The robot grips irregular objects without being affected by the gripping point. Abstract: Object parameters, like the mass and barycenter, are significant for the stability of object gripping. This study proposes a novel method to estimate the mass and barycenter coordinate of the task object based on the fingertip force/torque sensor. The equilibrium conditions of force and torque are deduced, and the estimation models are built for mass and barycenter coordinate. In particular, the principal component analysis (PCA) method is used to reduce inter-dimensional coupling and extract principal components for the multi-dimensional and multi-redundant signals of the fingertip force/torque sensor. Furthermore, an object parameter-based method is proposed for the task object gripping. Finally, we conduct a set of experiments to validate the effectiveness and generalization of the proposed method. The experiment results show that the parameter estimation method is effective to estimate the mass and barycenter coordinate, and the object parameter-based method can be generalized to grip irregular objects with varying gripping points.