Slippage and Deformation Preventive Control of Bionic Prosthetic Hands

Abstract

A proper grasping force is very difficult to obtain for a prosthetic hand due to a lack of detailed sensory feedback information. When the applied grasping force is too small, a heavy object may slip, whereas an overly large applied force may deform soft objects. This paper presents a bionic reflex control system designed to reconstruct the humanoid reflex control function for a prosthetic hand. First, a dynamical model of an underactuated tendon-driven prosthetic hand is obtained using the Lagrange method. Next, a force-sensing resistor sensor is adopted in the control loop to detect slippage based on empirical mode decomposition. Then, a polyvinylidene fluoride sensor is introduced into the control system to estimate the stiffness of an object, which is employed for modifying the initial desired grasping force to avoid deformation. The performance of the proposed bionic reflex control system is investigated through a series of experiments, the results of which successfully demonstrate its effectiveness, and verify that the grasping capability of a prosthetic hand is thereby improved.