Using Inverse Compensation Vectors for Autonomous Maze Exploration

Andrzej Bieszczad

2017

Abstract

Autonomous exploration of mazes requires finding a “center of gravity” keeping the robot safe from colliding with the walls. That is similar to the obstacle avoidance problem as the maze walls are obstacles that the robot must avoid. In this report, we describe an approach to controlling robot movements in a maze using an Inverse Compensation Vector (ICV) that is not much more computationally demanding than calculating a centroid point. The ICV is used to correct the robot velocity vector that determines the direction and the speed, so the robot moves in the maze staying securely within the passages between the walls. We have tested the approach using a simulator of a physical robot equipped with a planar LIDAR scanner. Our experiments showed that using the ICV to compensate robot velocity is an effective motion-correction method. Furthermore, we augmented the algorithm with preprocessing steps that alleviate problems caused by noisy raw data coming from actual LIDAR scans of a physical maze.

Download


Paper Citation


in Harvard Style

Bieszczad A. (2017). Using Inverse Compensation Vectors for Autonomous Maze Exploration . In Proceedings of the 14th International Conference on Informatics in Control, Automation and Robotics - Volume 2: ICINCO, ISBN 978-989-758-264-6, pages 397-404. DOI: 10.5220/0006437203970404


in Bibtex Style

@conference{icinco17,
author={Andrzej Bieszczad},
title={Using Inverse Compensation Vectors for Autonomous Maze Exploration},
booktitle={Proceedings of the 14th International Conference on Informatics in Control, Automation and Robotics - Volume 2: ICINCO,},
year={2017},
pages={397-404},
publisher={SciTePress},
organization={INSTICC},
doi={10.5220/0006437203970404},
isbn={978-989-758-264-6},
}


in EndNote Style

TY - CONF
JO - Proceedings of the 14th International Conference on Informatics in Control, Automation and Robotics - Volume 2: ICINCO,
TI - Using Inverse Compensation Vectors for Autonomous Maze Exploration
SN - 978-989-758-264-6
AU - Bieszczad A.
PY - 2017
SP - 397
EP - 404
DO - 10.5220/0006437203970404