Publication: Vertical line edge image-based fuzzy control of a two wheeled mobile robot
| dc.contributor.author | Neoh, Hock Seng | |
| dc.date.accessioned | 2024-01-19T02:26:16Z | |
| dc.date.available | 2024-01-19T02:26:16Z | |
| dc.date.issued | 2020-09-01 | |
| dc.description.abstract | Current research involving the use of edge images for fuzzy control of mobile robots are based on the utilization of steer angle magnitudes for actuation. This necessitates the fine tuning of the steer angle magnitudes to respond to certain feature points of its environment to perform a certain control action. This required training processes which could be time consuming. In this research, an edge image based fuzzy control system for a two wheeled mobile robot (2-WMR) which utilized the vertical line (VL) displacement was introduced. No fine tuning of the consequent sets was required. The steering control action was based on responding to the deviation of the VL in pixel counts. The standard deviation produced was 22-pixel counts from the center of the edge image. The tracks produced showed that the fuzzy control always directed the 2-WMR towards its landmark of interest in its various stages of movements. A fuzzy set with a threshold value determined the number of steps in each stage of movement. In addition, the fuzzy view descriptor was incorporated to provide the 2-WMR cognitive capabilities. The fuzzy view descriptor enabled the 2-WMR to ascertain its environment and select the appropriate VL at a certain stage of movement as a reference point to steer towards a landmark. A set of special registers determined the switching of the 2-WMR to consecutive view areas (VA). In this way the 2-WMR was able to steer from one landmark to another. Another novelty was the introduction of a new fuzzy edge detector to improve upon the conventional fuzzy edge detector. The new fuzzy edge detector demonstrated its improved robustness and efficiency to detect edges in various lighting conditions at various distances from the landmarks of interest. Since the new fuzzy edge detector maximizes the permutation of the masks in four direction, there was no need to further increment the fuzzy rules. This prevented the fuzzy rule explosion problem. | |
| dc.identifier.uri | https://erepo.usm.my/handle/123456789/18131 | |
| dc.language.iso | en | |
| dc.title | Vertical line edge image-based fuzzy control of a two wheeled mobile robot | |
| dc.type | Resource Types::text::thesis | |
| dspace.entity.type | Publication | |
| oairecerif.author.affiliation | Universiti Sains Malaysia |