Litcius/Paper detail

A new multi-objective algorithm for underactuated robotic finger during grasping and pinching assignments

Mortada A. Jabbar, Sadeq H. Bakhy, Enass H. Flaieh

2020IOP Conference Series Materials Science and Engineering12 citationsDOIOpen Access PDF

Abstract

Abstract This study presents new kinematic and kinetic equations of grasping and pinching assignments, which have been derived to consider the angles for the ternary solid links of a three-phalanx linkage underactuated robotic finger mechanism. Numerical and experimental investigations are conducted. Numerically, eight design criteria are adopted for finding an optimal solution using a new multi-objective function algorithm (grasping stability percentage, forces of grasping, squeezing force, grasping assignment mimic function, transmission angle for grasping assignment, mimic function for pinching assignment, transmission angle for pinching assignment and pinching force). The gradient descent method, comprising three steps, is used to find the optimal geometric parameters. These parameters are modified to fabricate the prototype of the robotic finger. Experimentally, the three-phalanx linkage finger is adjusted via the addition of a DC motor at the second linkage coupler link, having four bars for length adjustment for performing the aspect of human finger through the grasping and pinching assignments. The results show that adding ternary solid links to the robotic finger enhances the grasping stability and assists the complete embrace of objects, particularly objects that are smaller than those for the same finger without ternary solid links. In addition, the results indicate that the pinching force enhancement is 7–25% with reduction of the objects’ sizes, due to the ternary of solid links.

Topics & Concepts

Linkage (software)UnderactuationKinematicsReduction (mathematics)Function (biology)Ternary operationMechanism (biology)Computer scienceAlgorithmGRASPGradient descentControl theory (sociology)EngineeringSimulationArtificial intelligenceGeometryMathematicsPhysicsRobotClassical mechanicsBiochemistryBiologyQuantum mechanicsChemistryArtificial neural networkControl (management)GeneProgramming languageEvolutionary biologyRobot Manipulation and LearningSoft Robotics and ApplicationsRobotic Mechanisms and Dynamics