Skip to Main Content
 

Global Search Box

 
 
 
 

Files

File List


38052.pdf (14.65 MB)

ETD Abstract Container

Abstract Header

Cooperative Localization based Multi-Agent Coordination and Control

Chakraborty, Anusna
http://orcid.org/0000-0001-6292-3466
http://rave.ohiolink.edu/etdc/view?acc_num=ucin1613731906578577

Abstract Details

2021, PhD, University of Cincinnati, Engineering and Applied Science: Aerospace Engineering.
In this dissertation, the main focus is on developing a low-cost, robust, and efficient solution for Cooperative Localization to aid navigation of Unmanned Autonomous Vehicles in GPS-denied or degraded conditions. Initially, we derive conditions for complete observability of fixed-wing Unmanned Aerial Vehicles (UAVs) and Multi-rotor UAVs. A Relative Position Measurement Graph (RPMG) is created where the nodes of the graph are the vehicles or known features (landmarks) and the edges between them represent the measurements. Using graph theory and concepts of linear algebra, conditions for the maximum rank of the observability matrix are derived and a relationship between the rank of the observability matrix and the measurements available in the system are developed. One of the drawbacks of the conditions from this analysis is the necessity to maintain a connected RPMG at all time instants. Hence, a discrete-time observability condition is developed where the union of the RPMGs over a time interval has to be connected. Next, we address a fundamental problem for close coordination and control of Unmanned Vehicles (UVs). For various applications, the inertial position of the vehicles is not important. Relative pose and orientation among vehicles are useful for developing controllers in such cases. It is known that an Extended Kalman Filter (EKF) performs extremely well provided it is initialized close to the truth and receives measurements. For vehicles traveling long distances without any GPS measurements or with severe network delays such that they need to re-initialize the filters, the assumption of known a-priori is not valid. To circumvent these problems, a Multi-Hypothesis EKF (MHEKF) is developed with range-only measurements where the EKF has no a-priori information during initialization which means that the uncertainty associated is very large. In the end, we solve a distributed cooperative localization problem for ground vehicles. Centralized CL is computationally intensive. We develop a distributed cooperative localization algorithm such that every vehicle in the group estimates its own inertial state. This algorithm has been developed for Autonomous ground vehicles using range-only measurements in simulation.
Rajnikant Sharma, Ph.D. (Committee Chair)
Raj Bhatnagar, Ph.D. (Committee Member)
Kevin Brink, Ph.D. (Committee Member)
Kelly Cohen, Ph.D. (Committee Member)
Manish Kumar, Ph.D. (Committee Member)
175 p.
Robots
Unmanned Vehicles; Cooperation; Localization; Navigation; Relative; multi agent

Recommended Citations

Citations

  • Chakraborty, A. (2021). Cooperative Localization based Multi-Agent Coordination and Control [Doctoral dissertation, University of Cincinnati]. OhioLINK Electronic Theses and Dissertations Center. http://rave.ohiolink.edu/etdc/view?acc_num=ucin1613731906578577

    APA Style (7th edition)

  • Chakraborty, Anusna. Cooperative Localization based Multi-Agent Coordination and Control. 2021. University of Cincinnati, Doctoral dissertation. OhioLINK Electronic Theses and Dissertations Center, http://rave.ohiolink.edu/etdc/view?acc_num=ucin1613731906578577.

    MLA Style (8th edition)

  • Chakraborty, Anusna. "Cooperative Localization based Multi-Agent Coordination and Control." Doctoral dissertation, University of Cincinnati, 2021. http://rave.ohiolink.edu/etdc/view?acc_num=ucin1613731906578577

    Chicago Manual of Style (17th edition)

ucin1613731906578577
249
© 2020, all rights reserved.
This open access ETD is published by University of Cincinnati and OhioLINK.