Development of a neural network-based control system for the DLR-HIT II robot hand using leap motion

Abstract

In this paper, a neural network (NN) based adaptive controller has been successfully developed for the teleoperation of a DLR-HIT II robot hand using Leap Motion sensor. To achieve this, the coordinate positions of the human hand fingers are captured by a Leap Motion sensor. Moreover, inverse kinematics is used to transform the Cartesian position data of the fingertips into corresponding joint angles of all the five fingers, which are then directly sent to teleoperate the DLR-HIT II hand via User Datagram Protocol (UDP). In addition, a NN-based control system programmed by the MATLAB Simulink function has been investigated to enhance the overall control performance by compensating the dynamic uncertainties existing in the teleoperation system. The stability of the control system has been proved mathematically using Lyapunov stability equations. A series of experiments have been conducted to test the performance of the proposed control technique, which has been proved to be an effective teleoperation strategy for the DLR-HIT II robot hand.