Abstract:
Anthropomorphic robotic hands aim to resemble the functions and appearance of human hands. The increasing interest in their design arises from their importance in many medical and engineering applications. Controlling flexion and extension of each of the 4 fingers of an anthropomorphic robotic hand requires 3 motors to have full control of its joints. This study aims to construct a model for the 4 fingers to enable full control of all of their joints using only 2 motors by utilizing the correlation between proximal interphalangeal (PIP) and distal interphalangeal (DIP) joints of human finger to use a single motor for their control. A mechanical model is established for a single finger with correlated PIP and DIP joints to be generalized for the 4 fingers. Inverse kinematics is solved for the angles of the joints in terms of fingertip position. Transfer functions are derived for the change in tendons lengths during flexion to find the appropriate design of the pulleys controlling them. Hand parts are designed, 3D printed, and assembled accordingly, and then controlled using an Arduino microcontroller. Joints angles of the 4 fingers are measured over their full range of motion to calculate their correlation. Inverse kinematics gives a unique solution for each different fingertip position, but the ratio between the correlated angles should be assumed slightly different than that of a human hand for an algebraic solution to exist. The derived transfer functions are found to be nonlinear and increasing, meaning that the designed pulleys need to be spiral instead of perfectly circular. However, circular pulleys are used for the designed hand for simplification. Measured PIP and DIP joint angles show strong correlations for all fingers and over the full range of motion for all joints.|Keywords : Robotics, Anthropomorphic Hands, Interphalangeal Joints Correlation, Inverse Kinematics.
