Skip to Main Content
 

Global Search Box

 
 
 
 

Files

File List


miami1123877943.pdf (371.48 KB)

ETD Abstract Container

Abstract Header

LOCALIZATION VIA SINGLE AGENT PARTICLE FILTERS AND MULTI-AGENT BELIEF AGGREGATION IN ROBOCUP SOCCER

Thiagarajan, Priya
http://rave.ohiolink.edu/etdc/view?acc_num=miami1123877943

Abstract Details

2005, Master of Science, Miami University, Computer Science and Systems Analysis.
In a dynamic, partially observable, adversarial multi-agent domain such as Robocup soccer where the environment can change unpredictably due to actions of other agents, the need for agents to collaboratively maintain a consistent view of the world and to coherently merge exchanged information becomes crucial. We focus on implementing an aggregation-based multi-agent localization strategy for Robocup soccer agents. We extend the localization ability of UvA Trilearn, one of the top robotic soccer teams, to include particle filter aggregation [FBKT00] of the beliefs of individual agents. Aggregating agents’ beliefs about their relative positions enables them to improve their individual pose estimates as they detect each other. We assess our work with experiments that compare positioning error with and without aggregation. The results show that multi-agent belief aggregation does not have a major impact on agent localization in the RoboCup soccer domain. We provide detailed analysis of the reasons behind these results.
Pedrito Maynard-Zhang (Advisor)
53 p.
Computer Science

Recommended Citations

Citations

  • Thiagarajan, P. (2005). LOCALIZATION VIA SINGLE AGENT PARTICLE FILTERS AND MULTI-AGENT BELIEF AGGREGATION IN ROBOCUP SOCCER [Master's thesis, Miami University]. OhioLINK Electronic Theses and Dissertations Center. http://rave.ohiolink.edu/etdc/view?acc_num=miami1123877943

    APA Style (7th edition)

  • Thiagarajan, Priya. LOCALIZATION VIA SINGLE AGENT PARTICLE FILTERS AND MULTI-AGENT BELIEF AGGREGATION IN ROBOCUP SOCCER. 2005. Miami University, Master's thesis. OhioLINK Electronic Theses and Dissertations Center, http://rave.ohiolink.edu/etdc/view?acc_num=miami1123877943.

    MLA Style (8th edition)

  • Thiagarajan, Priya. "LOCALIZATION VIA SINGLE AGENT PARTICLE FILTERS AND MULTI-AGENT BELIEF AGGREGATION IN ROBOCUP SOCCER." Master's thesis, Miami University, 2005. http://rave.ohiolink.edu/etdc/view?acc_num=miami1123877943

    Chicago Manual of Style (17th edition)

miami1123877943
465
© 2005, all rights reserved.
This open access ETD is published by Miami University and OhioLINK.