End-Effector Impedance Control of Robotic Arm Based on Enhanced Neural Network RBF-PID-PSO

Abstract

The robotic arm's reliance on the location of the terminal based on camera positioning for electrical power transformer calibration systems has resulted in errors in the wiring terminals' insertion and extraction. Unfortunately, this approach fails to ensure exact assembly of transformer terminals. For a seamless terminal assembly, active compliance control must be incorporated between the wiring terminal and An end-effector on a robotic arm. This work offers an algorithm for control of mechanical arm end-effector impedance for terminal assembly according to RBF-PID-PSO. This algorithm modifies the center and weight matrix of the Radial Basis Function (RBF) neural network using a particle swarm optimization technique. The robotic arm model incorporates the impedance characteristics from the RBF neural network into an implementation of a PID algorithm that optimizes impedance model of end-effector forces. This makes it possible to convert end-effector forces into position errors in order to control impedance, which can subsequently be converted into six joint angles using the inverse kinematics of the robot. Simultaneously. The controller, demonstrated by Matlab experiments and simulations, is capable of control of end-effector forces through impedance, thus reducing errors in the control of force and torque. It also exhibits an improved ability to withstand force feedback signal disturbances, which greatly improves conforming control operation.