Design and development of mobile robot thematic mapping using flexible ellipse shape region

• Main Author: Mohd Hafiz, Taib
• Format: Thesis
• Language: English; English
• Published: 2014
• Subjects: T Technology (General); TJ Mechanical engineering and machinery
• Online Access: http://eprints.utem.edu.my/id/eprint/14916/1/DESIGN%20AND%20DEVELOPMENT%20OF%20MOBILE%20ROBOT%20THEMATIC%20MAPPING%20USING%20FLEXIBLE%20ELLIPSE%20SHAPE%20REGION%2024pages.pdf; http://eprints.utem.edu.my/id/eprint/14916/2/Design%20and%20development%20of%20mobile%20robot%20thematic%20mapping%20using%20flexible%20ellipse%20shape%20region.pdf; http://eprints.utem.edu.my/id/eprint/14916/; https://plh.utem.edu.my/cgi-bin/koha/opac-detail.pl?biblionumber=92054

Map is used to associate the entity of normal data distribution of an environment and also be used as a reference to detect changes in monitoring application. However, there is limitation in the use of a map if the designer of a robot needs to consider its resources, such as usage of memory space. The available map has a problem in terms of rigid structure or rigid perception of robot heading and indirectly uses a lot of memory space. Therefore, a new mapping technique called flexible ellipse shape region is proposed in this study. The ellipse boundary can be changed to accommodate normal data distribution of environment and it allows perception of robot heading to be mapped to normal data distribution from 0° until 360°. The objective of this study is to design and validate a new mapping technique, called flexible ellipse shape region. The performance of the map will be compared with grid map, perception based map and flexible region map in terms of memory space, access time and accuracy of map. Number of region is used to measure memory space of different maps. Meanwhile, the access time is calculated using time complexity, while accuracy of map is measured using new technique of confidence region. The experiments were conducted using Amigobot mobile robot in an L-shaped environment equipped with sonar sensor. The robot also has to carry a light sensor and a temperature sensor. The results of the experiments have shown that flexible ellipse shape region used 0.13%, 5%, 13.04% of memory space when being compared to grid map, perception-based map and flexible region map when being mapped with non-directional sensor data. In terms of access time, flexible ellipse shape region has used less time when being compared to perception based map and flexible region map. However, flexible ellipse shape region uses more access time when being compared to grid map. Lastly, map accuracy of flexible ellipse shape region is found to be higher, which is about 55.5% when being compared to flexible region map when being mapped with non-directional sensor data.