UMP Institutional Repository

Scan Matching and KNN Classification for Mobile Robot Localisation Algorithm

Addie, Irawan and Marni Azira, Markom and Abdul Hamid, Adom and Mohd Muslim Tan, E. S. (2017) Scan Matching and KNN Classification for Mobile Robot Localisation Algorithm. In: 3rd IEEE International Symposium on Robotic & Manufacturing Automation , 19-21 September 2017 , Universiti Putra Malaysia, Malaysia. . (Unpublished)

[img]
Preview
PDF
fkee-2017-addie-scan matching and knn1.pdf

Download (35kB) | Preview

Abstract

Mobile robots have made tremendous impact in our modern lives today, and its development is set to continue further. One of the most important domains to allow the interaction of mobile robots with human is its ability to know where it is in its environment, and how to navigate through it. This ability, however, needs algorithm has become more complex and requires high computational ability due to the demand for high accuracy, real time implementations and multi-tasking requirements. These are partly due to the need of multi-sensory system. This paper presents the use of single laser range finder for the mobile robot mapping and localisation system. The localisation algorithm is developed using scan matching method which is incorporated with K-nearest neighbours (KNN) classification. The mobile robot and the developed algorithm are tested in static environment. The results of the location estimation are able to achieve 80% of accuracy for each scan location with the distance range of ±2cm compared to the real location. As conclusion, the simple flow of the algorithm is suitable to replace the complex and high computational algorithm and system.

Item Type: Conference or Workshop Item (Lecture)
Uncontrolled Keywords: Mobile robot; KNN classification
Subjects: T Technology > TK Electrical engineering. Electronics Nuclear engineering
Faculty/Division: Faculty of Electrical & Electronic Engineering
Depositing User: Ir. Dr. Addie Irawan
Date Deposited: 31 Oct 2017 04:00
Last Modified: 31 Oct 2017 04:00
URI: http://umpir.ump.edu.my/id/eprint/18650
Download Statistic: View Download Statistics

Actions (login required)

View Item View Item