Aerial Robotic Systems and Methods for Tracking and Locating Multiple Mobile Objects with Range and Bearing Measuring Sensors

Simple item page
dc.contributor.advisorRanasinghe, Damith C.
dc.contributor.advisorRezatofighi, Hamid (Monash University)
dc.contributor.authorChen, Fei
dc.contributor.schoolSchool of Computer and Mathematics Sciences
dc.date.issued2024
dc.description.abstractUnmanned Aerial Vehicles (UAVs) capable of carrying various sensors have become versatile tools for tracking and monitoring applications. UAVs provide distinct advantages over traditional methods, such as flexibility, scalability, and the ability to cover large, remote areas while performing real-time monitoring. To realise this potential, UAVs must operate autonomously, making decisions in real-time to accomplish complex tasks. Developing UAV-based systems for tracking multiple mobile objects in dynamic, complex environments requires solutions for both multi-object tracking (MOT) and online path planning problems. MOT or tracking an unknown and time-varying number of objects while planning effective trajectories is a complex problem. The system must reliably track objects despite noisy, false, or missed measurements, while also planning efficient trajectories as multiple mobile objects appear and disappear within the search area. Practical considerations in UAV-based tracking—such as limited sensor capabilities, computational power, and communication bandwidth add to the difficulty of realising an autonomous system. Additionally, application constraints, such as maintaining a safe distance from objects of interest, further increase the complexity of the problem. This dissertation’s objective is it to investigate these theoretical and practical challenges to develop robust, efficient, fully autonomous UAV systems for tracking multiple mobile objects under real-world conditions. To this end, an aerial robot for locating radio-tagged wildlife in challenging terrains is developed first. The system leverages the unique characteristics of received signal strength and rotation-for-bearing measurements to improve tracking performance and reduce flight-times. Second, a pseudo-bearings for radio source localisation is developed. The approach utilises the flight dynamics of multirotor UAVs to realise faster pseudo-bearing measurements instead of the previous, slow, rotation-for-bearing method. Third, a track-before-detect filter for a Pseudo-Doppler direction-finding system is developed, offering a robust solution for tracking fast-moving radio sources in noisy environments. Fourth, a fast, distributed multi-object tracking algorithm is formulated for a network of heterogeneous sensors with unknown fields of view (FoVs). The method overcomes the limitations of using a single sensor with limited FoV for tracking and is significantly more computationally efficient than state-of-the-art methods but achieves comparable tracking accuracy and bandwidth efficiency.
dc.description.dissertationThesis (Ph.D.) -- University of Adelaide, School of Computer and Mathematics Sciences, 2025en
dc.identifier.urihttps://hdl.handle.net/2440/146400
dc.language.isoen
dc.provenanceThis electronic version is made publicly available by the University of Adelaide in accordance with its open access policy for student theses. Copyright in this thesis remains with the author. This thesis may incorporate third party material which has been used by the author pursuant to Fair Dealing exceptions. If you are the owner of any included third party copyright material you wish to be removed from this electronic version, please complete the take down form located at: http://www.adelaide.edu.au/legalsen
dc.subjectLocalization
dc.subjectAerial System
dc.subjectMulti-object Tracking
dc.subjectRadio Sources
dc.subjectPath Planning
dc.subjectField Robotics
dc.titleAerial Robotic Systems and Methods for Tracking and Locating Multiple Mobile Objects with Range and Bearing Measuring Sensors
dc.typeThesisen

Files

Original bundle
Now showing 1 - 1 of 1
No Thumbnail Available
Name:
ChenF2024_PhD.pdf
Size:
30.17 MB
Format:
Adobe Portable Document Format
Download

Collections

Research Theses