PROJECT TITLE :

Adaptive Neural Network Control of a Compact Bionic Handling Arm

ABSTRACT:

In this paper, autonomous management problem of a class of bionic continuum robots named “Compact Bionic Handling Arm” (CBHA) is addressed. These robots will reproduce biological behaviors of trunks, tentacles, or snakes. The modeling problem related to continuum robots includes nonlinearities, structured and unstructured uncertainties, and therefore the hyperredundancy. Moreover to those problems, the CBHA comprises the hysteresis behavior of its actuators and a memory phenomenon related to its structure made of polyamide materials. These undesirable effects build it difficult to design a control system primarily based on quantitative models of the CBHA. So, 2 subcontrollers are proposed in this paper. One, encapsulated in the opposite, and each implemented in real time permit controlling of the CBHA's finish-effector position. The first subcontroller controls the CBHA's kinematics based mostly on a distal supervised learning theme. The second subcontroller controls the CBHA's kinetics primarily based on an adaptive neural management. These subcontrollers allow a higher assessment of the stability of the control architecture while making certain the convergence of Cartesian errors. The obtained experimental results employing a CBHA robot show an accurate tracking of the CBHA's end-effector position.