An overall grasp evaluation function to assess and optimize prosthetic hands

Abstract

In this study, we create an evolution function that assesses prosthetic hands in terms of their ability to grasp various objects like a human hand. It is required to evaluate the reachable workspace of prosthetic hands and grasp ability to assess it completely. The first one can be done by using forward kinematics. The Volume of the Grasp Wrench space, one of the grasp qualities indexes, is used to measure the second one. Randomization and grasp taxonomy are used to generalize the grasp quality to indicate the prosthetic hand functionality. Afterward, the created evaluation function is used to specify the importance of each finger and DOF. the results show that the most significant finger is the Thumb. The most crucial DOFs are the abduction movement of the Thumb's CMC joint and the Index's MCP joint. Finally, optimized hand configuration is created, considering limitations which are a number of actuators and stationery DOFs. Optimization has two steps. In the first step, the stationary joints and the joints that should be actuated independently from each other are chosen. In the second step, the remaining DOFs are specified to be coupled with which actuator using the Taguchi method.