Sistema estabilizador da adesão de um robô escalador com rodas magnéticas

Abstract

This work consists in the design of an adhesion stabilization system of a climbing robot with magnetic wheels. The robot’s design comes from the need to automatize the inspection process of industrial storage tanks for petroleum products, which is currently performed manually. The robot is being developed in the Laboratory of Automation and Advanced Control Systems (LASCA) of the Federal Technological University of Paraná (UTFPR). First, a complete a theoretical analysis of the prototype was carried out including a study of the robot’s structure, its requirements and the kinematics and dynamics analyses. Then, a study of the robot’s magnetic wheels and the characteristics of the magnetic field in the occurancy of detachment between the magnetic wheel and the ferro-magnetic surfaces was carried out. The magnetic field data is acquired through the magnetometer of the inertial measurement unit sensor of the robot. Then a multilayer perceptron artificial neural network was implemented in order to interpret the magnetic field data and estimate the adhesion force between robot and surface. Finally the adhesion force quantification is used to implement an adhesion control system for the robot.