Title: Coordination control and obstacle avoidance for a team of mobile robots in unknown environment

Authors: Sucheta Roy; Md Assad-Uz-Zaman; Mohammad Habibur Rahman; Md Suruz Miah

Addresses: Mechanical Engineering Department, University of Wisconsin-Milwaukee, Milwaukee, 53211, USA ' Mechanical Engineering Department, University of Wisconsin-Milwaukee, Milwaukee, 53211, USA ' Mechanical Engineering Department, University of Wisconsin-Milwaukee, Milwaukee, 53211, USA ' Department of Electrical and Computer Engineering, Bradley University, Peoria, 61625, USA

Abstract: This paper presents a dual-level control structure for controlling a mobile robot and/or a team of mobile robots to navigate through an unknown (static or dynamic) environment. The higher-level controller operates in cooperation with robot's state estimation and mapping algorithm, extended Kalman filter - simultaneous localisation and mapping (EKF-SLAM), and the lower-level controller (PID) controls the motion of the robot when it encounters an obstacle and manoeuvres the robot around the obstacle/s. The higher-level controller jumps in as soon as the robot is out of the obstacle range and moves the robot to the goal. The obstacle avoidance technique involves a novel approach to calculate the rebound angle. The experimental results show that the proposed dual-level control structure can be effectively used to manoeuvre a mobile robot or a team of mobile robots to navigate through a dynamic environment.

Keywords: EKF-SLAM; extended kalman filter - simultaneous localisation and mapping; mobile robots; multi-robot systems; dual-level motion controller; obstacle avoidance; PID control; autonomous robots; omni-directional robots; leader-follower.

DOI: 10.1504/IJMIC.2020.112296

International Journal of Modelling, Identification and Control, 2020 Vol.34 No.4, pp.316 - 327

Received: 22 Jan 2020
Accepted: 08 Mar 2020
Published online: 07 Jan 2021 *

Full-text access for editors Full-text access for subscribers Purchase this article Comment on this article