Articulated robots motion planning using foraging ant strategy
Many different approaches to tackle the problem of motion planning for articulated robots in an environment with obstacles based on random sampling have been proposed. One popular approach is called single-query bi-directional motion planning with a lazy collision checking probabilistic road map (SB...
| Main Author: | |
|---|---|
| Format: | Article |
| Language: | English |
| Published: |
Penerbit UTM Press
2008
|
| Subjects: | |
| Online Access: | http://eprints.utm.my/9298/3/MohdMurtadhaMohamad2008_ArticulatedRobotsMotionPlanningUsingForaging.pdf |
| _version_ | 1825910301050011648 |
|---|---|
| author | Mohamad, Mohd. Murtadha |
| author_facet | Mohamad, Mohd. Murtadha |
| author_sort | Mohamad, Mohd. Murtadha |
| collection | ePrints |
| description | Many different approaches to tackle the problem of motion planning for articulated robots in an environment with obstacles based on random sampling have been proposed. One popular approach is called single-query bi-directional motion planning with a lazy collision checking probabilistic road map (SBL-PRM). However, the performance of this method is sub-optimal in terms of the number of configurations generated, length of path, amount of collision checking and computational time. To improve the performance, those aspects must be considered further as they are inter-related with each other. A novel modification of SBLPRM that decreases the size of excessive configurations in the roadmap, by incrementally building a one-tree structure originating from the start configuration, is presented. This approach, the single-query unidirectional approach with lazy collision checking (SUL-PRM), has experimentally shown to be equal to the SBL-PRM. However, there still exists generated configurations that were excluded from the successful path. The generation of these unconsumed configurations corresponding to the tree structure has pointlessly utilized the computational resources and affected the planning time. Hence, a new method of configuration generation along with a novel searching style is devised. An alternative search approach using ant behaviour in a robotics application is applied. This paper proposes a novel search technique, the F-Ant algorithm, in order to find a reliable path between the initial configuration and the goal configuration of the articulated robot. This novel algorithm, taking two input configurations, explores the robot's free space by building up a unidirectional search beginning at the initial configuration. The planner samples the free configuration repetitively in the neighbourhood within the radius of the current configuration, and tests the edge for a collision-free path between the new sampled configurations, until it is connected to the goal configuration. Simulation and experimental comparisons of F-Ant and SBL-PRM have been conducted, showing the performance differences between these two techniques. |
| first_indexed | 2024-03-05T18:15:03Z |
| format | Article |
| id | utm.eprints-9298 |
| institution | Universiti Teknologi Malaysia - ePrints |
| language | English |
| last_indexed | 2024-03-05T18:15:03Z |
| publishDate | 2008 |
| publisher | Penerbit UTM Press |
| record_format | dspace |
| spelling | utm.eprints-92982017-11-01T04:17:22Z http://eprints.utm.my/9298/ Articulated robots motion planning using foraging ant strategy Mohamad, Mohd. Murtadha TJ Mechanical engineering and machinery QA75 Electronic computers. Computer science Many different approaches to tackle the problem of motion planning for articulated robots in an environment with obstacles based on random sampling have been proposed. One popular approach is called single-query bi-directional motion planning with a lazy collision checking probabilistic road map (SBL-PRM). However, the performance of this method is sub-optimal in terms of the number of configurations generated, length of path, amount of collision checking and computational time. To improve the performance, those aspects must be considered further as they are inter-related with each other. A novel modification of SBLPRM that decreases the size of excessive configurations in the roadmap, by incrementally building a one-tree structure originating from the start configuration, is presented. This approach, the single-query unidirectional approach with lazy collision checking (SUL-PRM), has experimentally shown to be equal to the SBL-PRM. However, there still exists generated configurations that were excluded from the successful path. The generation of these unconsumed configurations corresponding to the tree structure has pointlessly utilized the computational resources and affected the planning time. Hence, a new method of configuration generation along with a novel searching style is devised. An alternative search approach using ant behaviour in a robotics application is applied. This paper proposes a novel search technique, the F-Ant algorithm, in order to find a reliable path between the initial configuration and the goal configuration of the articulated robot. This novel algorithm, taking two input configurations, explores the robot's free space by building up a unidirectional search beginning at the initial configuration. The planner samples the free configuration repetitively in the neighbourhood within the radius of the current configuration, and tests the edge for a collision-free path between the new sampled configurations, until it is connected to the goal configuration. Simulation and experimental comparisons of F-Ant and SBL-PRM have been conducted, showing the performance differences between these two techniques. Penerbit UTM Press 2008-12 Article PeerReviewed application/pdf en http://eprints.utm.my/9298/3/MohdMurtadhaMohamad2008_ArticulatedRobotsMotionPlanningUsingForaging.pdf Mohamad, Mohd. Murtadha (2008) Articulated robots motion planning using foraging ant strategy. Jurnal Teknologi Maklumat, 20 (4). pp. 163-181. ISSN 0128-3790 |
| spellingShingle | TJ Mechanical engineering and machinery QA75 Electronic computers. Computer science Mohamad, Mohd. Murtadha Articulated robots motion planning using foraging ant strategy |
| title | Articulated robots motion planning using foraging ant strategy |
| title_full | Articulated robots motion planning using foraging ant strategy |
| title_fullStr | Articulated robots motion planning using foraging ant strategy |
| title_full_unstemmed | Articulated robots motion planning using foraging ant strategy |
| title_short | Articulated robots motion planning using foraging ant strategy |
| title_sort | articulated robots motion planning using foraging ant strategy |
| topic | TJ Mechanical engineering and machinery QA75 Electronic computers. Computer science |
| url | http://eprints.utm.my/9298/3/MohdMurtadhaMohamad2008_ArticulatedRobotsMotionPlanningUsingForaging.pdf |
| work_keys_str_mv | AT mohamadmohdmurtadha articulatedrobotsmotionplanningusingforagingantstrategy |