Abstract—The hyper-redundant robot can be used in object
manipulation task. Due to the complexity of the mechanism,
the object manipulation task can be reduced into encircling the
robot around an object. The robot is required to be formed
into some shapes to cage around an object and then moves an
object. This method presented in this paper combine the caging
process which is the shape control solution for hyperredundant arm using the virtual constraint together with the
virtual push force and the manipulating process which is the
inverse kinematics solution for hyper-redundant arm using the
Neural Network and Bezier curve together with the gradient
descent method. The algorithm allows the robot to be able to
encircle and move the object to the desired position without
grasping. A computer simulation of the serial link manipulator
has demonstrated the effectiveness of the proposed method.
Index Terms—Hyper-redundant robot, neural network, bezier curve, gradient descent method
C. Jareanpon is with the Department of Computer Science, Faculty of Informatics, Mahasarakham University, Thailand (e-mail: chatklaw.j@gmail.com).
[PDF]
Cite: Chatklaw Jareanpon, "Encircled Hyper-Redundant Manipulation Using Virtual Constraint, Bezier Curve and Gradient Descent Method,"
International Journal of Future Computer and Communication vol. 1, no. 2, pp. 101-105, 2012.
