Control of mobile robots using ActionLib

Abstract

Mobile robots are very complex systems and involve the integration of various structures (mechanical, software and electronics). The robot control system must integrate these structures so that it can perform its tasks properly. Mobile robots use control strategies for many reasons, like velocity control of wheels, position control and path tracking. These controllers require the use of preemptive structures. Therefore, this tutorial chapter aims to clarify the design of controllers for mobile robots based on ROS ActionLib. Each controller is designed in an individual ROS node to allow parallel processing by operating system. To exemplify the controller design using ActionLib, this chapter will demonstrate the implementation of two different types of controllers (PID and Fuzzy) for position control of a servo motor. These controllers will be available on GitHub. Also, a case study of scheduled fuzzy controllers based on ROS ActionLib for a magnetic climber robot used in the inspection of spherical tanks will be shown.