Improved Range-Only Beacon Initialization Towards Localization System
Molaletsa Namoshe1, Oduetse Matsebe2, Ngatho Tlale3
1Molaletsa Namoshe, College of Engineering, Department of Mechanical & Energy Engineering, Botswana International University of Science & Technology, Palapye, Botswana.
2Oduetse Matsebe, College of Engineering, Department of Mechanical & Energy Engineering, Botswana International University of Science & Technology, Palapye, Botswana.
3Nkgato Tlale, Transnet, Carlton Centre, 150 Commissioner Street, Johannesburg, South Africa.
Manuscript received on 15 June 2015 | Revised Manuscript received on 25 June 2015 | Manuscript Published on 30 June 2015 | PP: 38-41 | Volume-4 Issue-5, June 2015 | Retrieval Number: E4076064515/15©BEIESP
Open Access | Editorial and Publishing Policies | Cite | Mendeley | Indexing and Abstracting
© The Authors. Blue Eyes Intelligence Engineering and Sciences Publication (BEIESP). This is an open access article under the CC-BY-NC-ND license (http://creativecommons.org/licenses/by-nc-nd/4.0/)
Abstract: Mobile robot operation in an un-surveyed environment presents a challenging problem, particularly in GPS denied spaces. The complexity of the problem scales up if the sensor used to aid navigation can only provide range information about the features in that environment. In the past, almost all solutions to Localization problems relied on a prior knowledge of feature locations. In this paper however, range measurements, characteristically known to have outliers and unobservable are used to solve the localization problem. Past approaches to this problem have used delayed initialization of newly observed feature(s) until good estimates are available; a process akin to Hough transforms methods. This ratio thresholding approach has shown to be susceptible to system divergence, especially when large environments are explored. In this paper therefore, a pose disambiguating algorithms comprising of outlier rejection, particle swam optimization (PSO) and an area under a probability distribution function (pdf) methods are used to solve the localization system using real data acquired by a mobile robot in an unknown space. To validate the proposed methods, experimental real data sets obtain by Odyssey III during the GOATS’02 experiments are used.
Keywords: Range Data, Gaussian Distribution, Localization, Feature Initialization, Beacon (Feature/ Landmark), and Observation Sensor.
Scope of the Article: WSN