國立虎尾科技大學 |

Distributed ,and Certifiable Optimization for Multi-Robot Systems.

Record Type:	Language materials, manuscript : Monograph/item
Title/Author:	Distributed ,and Certifiable Optimization for Multi-Robot Systems./
Author:	Halsted, Trevor Neill.
Description:	1 online resource (180 pages)
Notes:	Source: Dissertations Abstracts International, Volume: 85-11, Section: B.
Contained By:	Dissertations Abstracts International85-11B.
Subject:	Robots. -
Online resource:	click for full text (PQDT)
ISBN:	9798382637587

Distributed ,and Certifiable Optimization for Multi-Robot Systems.
Halsted, Trevor Neill.

Distributed ,and Certifiable Optimization for Multi-Robot Systems. - 1 online resource (180 pages)

Source: Dissertations Abstracts International, Volume: 85-11, Section: B.

Thesis (Ph.D.)--Stanford University, 2023.

Includes bibliographical references

This thesis addresses two key challenges in applying optimization methods to multi-robot systems. First, the need for Distributed Optimization arises in unstructured and remote environments in which centralized coordination is impractical or infeasible. Second, the nonconvexity of many relevant problems necessitates certifiable optimization, which provides provable bounds on the quality of a solution. We present new techniques for both distributed optimization and certifiable optimization, as well as a unified framework for distributed and certifiable multi-robot optimization.Several tools presented in this thesis enable robots to achieve system-wide coordination through pairwise communication. The ability to cooperatively solve optimization problems through distributed robot-to-robot communication is a crucial component of multi-robot autonomy. We introduce a communication-efficient general-purpose distributed optimization method, Separable Optimization Variable ADMM, and demonstrate its scalability and robustness. Additionally, an optimization-based approach to distributed estimation replicates the performance of the centralized Kalman Filter. We prove that our Distributed Rolling-Window Tracking algorithm maintains consistent estimates without communicating covariance matrices as alternative methods require.Our approach to certifiable optimization considers a broad class of synchronization problems that admit a semidefinite relaxation. We present a novel reformulation of the distance-based localization problem, showing for the first time that it belongs to this same class of certifiable optimization problems. Compared to other distance-based localization algorithms, our method recovers good solutions over a range of problem instantiations.Finally, we present a new approach to distributed certifiable optimization. Our technique for solving a decomposition of a sparse semidefinite program introduces a novel "dualized agreement" constraint. This formulation allows networked robots to iteratively solve for their individual components in a large semidefinite program. Together, these contributions provide a general framework for fully distributed and provably optimalmulti-robot coordination.

Electronic reproduction.
Ann Arbor, Mich. :
ProQuest,
2024

Mode of access: World Wide Web

ISBN: 9798382637587Subjects--Topical Terms:

654842
Robots.
Index Terms--Genre/Form:

554714
Electronic books.

Distributed ,and Certifiable Optimization for Multi-Robot Systems.
LDR:03487ntm a22003617 4500 001 1151210
005 20241104055828.5
006 m o d
007 cr bn ---uuuuu
008 250605s2023 xx obm 000 0 eng d
020 $a 9798382637587
035 $a (MiAaPQ)AAI31049694
035 $a (MiAaPQ)STANFORDvp445pw4013
035 $a AAI31049694
040 $a MiAaPQ $b eng $c MiAaPQ $d NTU
100 1 $a Halsted, Trevor Neill. $3 1477934
245 1 0 $a Distributed ,and Certifiable Optimization for Multi-Robot Systems.
264 0 $c 2023
300 $a 1 online resource (180 pages)
336 $a text $b txt $2 rdacontent
337 $a computer $b c $2 rdamedia
338 $a online resource $b cr $2 rdacarrier
500 $a Source: Dissertations Abstracts International, Volume: 85-11, Section: B.
500 $a Advisor: Schwager, Mac.
502 $a Thesis (Ph.D.)--Stanford University, 2023.
504 $a Includes bibliographical references
520 $a This thesis addresses two key challenges in applying optimization methods to multi-robot systems. First, the need for Distributed Optimization arises in unstructured and remote environments in which centralized coordination is impractical or infeasible. Second, the nonconvexity of many relevant problems necessitates certifiable optimization, which provides provable bounds on the quality of a solution. We present new techniques for both distributed optimization and certifiable optimization, as well as a unified framework for distributed and certifiable multi-robot optimization.Several tools presented in this thesis enable robots to achieve system-wide coordination through pairwise communication. The ability to cooperatively solve optimization problems through distributed robot-to-robot communication is a crucial component of multi-robot autonomy. We introduce a communication-efficient general-purpose distributed optimization method, Separable Optimization Variable ADMM, and demonstrate its scalability and robustness. Additionally, an optimization-based approach to distributed estimation replicates the performance of the centralized Kalman Filter. We prove that our Distributed Rolling-Window Tracking algorithm maintains consistent estimates without communicating covariance matrices as alternative methods require.Our approach to certifiable optimization considers a broad class of synchronization problems that admit a semidefinite relaxation. We present a novel reformulation of the distance-based localization problem, showing for the first time that it belongs to this same class of certifiable optimization problems. Compared to other distance-based localization algorithms, our method recovers good solutions over a range of problem instantiations.Finally, we present a new approach to distributed certifiable optimization. Our technique for solving a decomposition of a sparse semidefinite program introduces a novel "dualized agreement" constraint. This formulation allows networked robots to iteratively solve for their individual components in a large semidefinite program. Together, these contributions provide a general framework for fully distributed and provably optimalmulti-robot coordination.
533 $a Electronic reproduction. $b Ann Arbor, Mich. : $c ProQuest, $d 2024
538 $a Mode of access: World Wide Web
650 4 $a Robots. $3 654842
650 4 $a Decomposition. $3 1372547
650 4 $a Automobiles. $3 1123680
650 4 $a Convex analysis. $3 1477935
650 4 $a Semidefinite programming. $3 890223
650 4 $a Communication. $3 556422
650 4 $a Optimization techniques. $3 1413598
650 4 $a Autonomous vehicles. $3 981632
650 4 $a Mathematics. $3 527692
650 4 $a Robotics. $3 561941
650 4 $a Transportation. $3 558117
655 7 $a Electronic books. $2 local $3 554714
690 $a 0459
690 $a 0405
690 $a 0771
690 $a 0709
710 2 $a ProQuest Information and Learning Co. $3 1178819
710 2 $a Stanford University. $3 1184533
773 0 $t Dissertations Abstracts International $g 85-11B.
856 4 0 $u http://pqdd.sinica.edu.tw/twdaoapp/servlet/advanced?query=31049694 $z click for full text (PQDT)