Robot Arm Design for Children Writing Ability Enhancement using Cartesian Equations based on ANFIS

Abstract

Writing is the main part of children's advancement. This paper suggests a real robotic methodology for teaching children how to write and enhance their writing and drawing ability in general. This method includes design with the implementation of two DOF (Degrees Of Freedom) robot arm. The robot trajectory planning motion was done using multi-segment parametric Cartesian equations based on the Adaptive Neuro-Fuzzy Inference System (ANFIS) for modeling the inverse kinematics to PWM directly. The proposed overall structure contains two ANFIS structures. The inputs of each ANFIS are the desired Cartesian coordination represents the desired letter points while the outputs are the two Pulse Width Modulation (PWM) servomotor commands needed to actuate each robot link to the desired position. The mechanical structure of the robot arm uses three servo motors. The last servomotor used for raising and lowering the pen, which is attached mechanically to the robot end-effector. In this work, the maximum position error of robot end-effector is evaluated between theoretical and experimental work. These position errors in X-axis do not exceed (±0.0170 m) and in Y-axis not exceed (±0.0150 m). The results of position errors are acceptable and occur due to the commercial servomotors used. The written letters are clear, smooth and always inside the robot reachable area.