Relative localization based on the fusion of ultra-wideband and LiDAR in robot swarms
Multi-robot technology, as a major research hot spot in the field of robotics, has attracted widespread attention. Its core advantage lies in its unique cooperation, enabling it to be widely applied in many scenarios and fields. In particular, decentralized robot swarms, with their outstanding flexi...
| Main Author: | |
|---|---|
| Other Authors: | |
| Format: | Thesis-Master by Coursework |
| Language: | English |
| Published: |
Nanyang Technological University
2024
|
| Subjects: | |
| Online Access: | https://hdl.handle.net/10356/177892 |
| _version_ | 1826116681341075456 |
|---|---|
| author | Wu, Yunbin |
| author2 | Chau Yuen |
| author_facet | Chau Yuen Wu, Yunbin |
| author_sort | Wu, Yunbin |
| collection | NTU |
| description | Multi-robot technology, as a major research hot spot in the field of robotics, has attracted widespread attention. Its core advantage lies in its unique cooperation, enabling it to be widely applied in many scenarios and fields. In particular, decentralized robot swarms, with their outstanding flexibility and autonomy, as well as no requirements for central server, allow them to collaborate on a large scale, thus becoming a focus of research. This study proposes a new algorithm that combines Ultra-WideBand (UWB) and LiDAR technologies for the recognition, relative localization, and tracking of nearby peer robots. The requirement for robot hardware is low, making it suitable for non-Simultaneous Localization and Mapping (SLAM) robots. The algorithm utilizes LiDAR-collected point cloud data and UWB’s distance information for environmental perception, clustering, and tracking, successfully distinguishing robots from environmental obstacles and tracking the movement of peer robots in real-time. By adopting the Kalman filter to stably track clustered targets, and integrating distance-based methods and historical tracks matching techniques to complete the tracking task, the experimental results show that the algorithm achieves a high recognition rate of 90.02% and a MAE of 0.0491m, which is lower than the robot’s outer diameter size, ensuring that the clustering center remains stable inside the robot’s outline. The visualization results illustrate that the algorithm can effectively track and distinguish the positions of peer robots throughout the entire process, demonstrating a high degree of reliability, stability, providing valuable reference for relative localization in robot swarms. |
| first_indexed | 2024-10-01T04:15:21Z |
| format | Thesis-Master by Coursework |
| id | ntu-10356/177892 |
| institution | Nanyang Technological University |
| language | English |
| last_indexed | 2024-10-01T04:15:21Z |
| publishDate | 2024 |
| publisher | Nanyang Technological University |
| record_format | dspace |
| spelling | ntu-10356/1778922024-06-07T15:43:33Z Relative localization based on the fusion of ultra-wideband and LiDAR in robot swarms Wu, Yunbin Chau Yuen School of Electrical and Electronic Engineering chau.yuen@ntu.edu.sg Engineering Robot swarms Decentralized Peers' localization UWB LiDAR Multi-robot technology, as a major research hot spot in the field of robotics, has attracted widespread attention. Its core advantage lies in its unique cooperation, enabling it to be widely applied in many scenarios and fields. In particular, decentralized robot swarms, with their outstanding flexibility and autonomy, as well as no requirements for central server, allow them to collaborate on a large scale, thus becoming a focus of research. This study proposes a new algorithm that combines Ultra-WideBand (UWB) and LiDAR technologies for the recognition, relative localization, and tracking of nearby peer robots. The requirement for robot hardware is low, making it suitable for non-Simultaneous Localization and Mapping (SLAM) robots. The algorithm utilizes LiDAR-collected point cloud data and UWB’s distance information for environmental perception, clustering, and tracking, successfully distinguishing robots from environmental obstacles and tracking the movement of peer robots in real-time. By adopting the Kalman filter to stably track clustered targets, and integrating distance-based methods and historical tracks matching techniques to complete the tracking task, the experimental results show that the algorithm achieves a high recognition rate of 90.02% and a MAE of 0.0491m, which is lower than the robot’s outer diameter size, ensuring that the clustering center remains stable inside the robot’s outline. The visualization results illustrate that the algorithm can effectively track and distinguish the positions of peer robots throughout the entire process, demonstrating a high degree of reliability, stability, providing valuable reference for relative localization in robot swarms. Master's degree 2024-06-03T05:11:25Z 2024-06-03T05:11:25Z 2024 Thesis-Master by Coursework Wu, Y. (2024). Relative localization based on the fusion of ultra-wideband and LiDAR in robot swarms. Master's thesis, Nanyang Technological University, Singapore. https://hdl.handle.net/10356/177892 https://hdl.handle.net/10356/177892 en application/pdf Nanyang Technological University |
| spellingShingle | Engineering Robot swarms Decentralized Peers' localization UWB LiDAR Wu, Yunbin Relative localization based on the fusion of ultra-wideband and LiDAR in robot swarms |
| title | Relative localization based on the fusion of ultra-wideband and LiDAR in robot swarms |
| title_full | Relative localization based on the fusion of ultra-wideband and LiDAR in robot swarms |
| title_fullStr | Relative localization based on the fusion of ultra-wideband and LiDAR in robot swarms |
| title_full_unstemmed | Relative localization based on the fusion of ultra-wideband and LiDAR in robot swarms |
| title_short | Relative localization based on the fusion of ultra-wideband and LiDAR in robot swarms |
| title_sort | relative localization based on the fusion of ultra wideband and lidar in robot swarms |
| topic | Engineering Robot swarms Decentralized Peers' localization UWB LiDAR |
| url | https://hdl.handle.net/10356/177892 |
| work_keys_str_mv | AT wuyunbin relativelocalizationbasedonthefusionofultrawidebandandlidarinrobotswarms |