東華大學圖書館 |

Language: English

Help

回圖書館首頁

手機版館藏查詢

Back

Switch To: Labeled | MARC Mode | ISBD

Addressing Stiffness-Induced Challen...

Halm, Mathew.

Linked to FindBook

Google Book

Amazon

博客來

Addressing Stiffness-Induced Challenges in Modeling and Identification for Rigid-Body Systems With Friction and Impacts.

Record Type:	Electronic resources : Monograph/item
Title/Author:	Addressing Stiffness-Induced Challenges in Modeling and Identification for Rigid-Body Systems With Friction and Impacts./
Author:	Halm, Mathew.
Published:	Ann Arbor : ProQuest Dissertations & Theses, : 2023,
Description:	174 p.
Notes:	Source: Dissertations Abstracts International, Volume: 85-03, Section: B.
Contained By:	Dissertations Abstracts International85-03B.
Subject:	Robotics. -
Online resource:	https://pqdd.sinica.edu.tw/twdaoapp/servlet/advanced?query=30568375
ISBN:	9798380384766

Addressing Stiffness-Induced Challenges in Modeling and Identification for Rigid-Body Systems With Friction and Impacts.
Halm, Mathew.

Addressing Stiffness-Induced Challenges in Modeling and Identification for Rigid-Body Systems With Friction and Impacts. - Ann Arbor : ProQuest Dissertations & Theses, 2023 - 174 p.

Source: Dissertations Abstracts International, Volume: 85-03, Section: B.

Thesis (Ph.D.)--University of Pennsylvania, 2023.

Imperfect, useful dynamical models have enabled significant progress in planning and controlling robotic locomotion and manipulation. Traditionally, these models have been physics-based, with accuracy relying upon manual calibration only feasible in laboratory environments. As robotics expands into complex real-world applications, models of unknown environments must instead be automatically fit to limited data. One major challenge is modeling frictional contact, especially during collisions involved in common robotics tasks. Rapid deformation under impact manifests as extreme sensitivity to initial conditions and material properties. Thus, even slight errors in state estimation and system identification can lead to significant prediction errors. Consequently, model inaccuracy or the sim-to-real gap often hinders the development of performant robotics algorithms.Physical models can be optimized using advanced techniques to overcome these challenges, but such methods have limited tractability when a large number of parameters are unknown. Furthermore, even given accurate parameters, roboticists often make inaccurate rigid-body approximations to reduce the computational burdens of physical simulation to meet faster-than-real-time requirements. An alternative black-box approach, in which models are learned from scratch, has attempted to address these issues for instance using deep neural networks (DNN's). While DNNs in theory can capture any dynamical behavior, they empirically struggle with the stiff behaviors associated with contact.This dissertation instead focuses on scaling physical model identification to the high-dimensional setting and quantifying the limited accuracy of low-fidelity models. We consider rigid bodies under-going rigid contact, for which infinite stiffness is represented as constrained optimization. By careful treatment of these constraints, we demonstrate that infinitely-stiff dynamics can be identified by optimizing a non-stiff objective. In conjunction, we use DNN's in a white-box setting to model physical quantities, specifically reconstructing geometries from scratch. We then consider how rigid-body collision models lack the fidelity to correctly predict outcomes of nearly-simultaneous impacts-such as heel and toe strikes during a footstep. We develop a theoretical basis to capture partial knowledge of impact events as uncertain set-valued outcomes, and again use numerical optimization to compute approximations of such sets.

ISBN: 9798380384766Subjects--Topical Terms:

519753
Robotics.
Subjects--Index Terms:

Robotics algorithms

Addressing Stiffness-Induced Challenges in Modeling and Identification for Rigid-Body Systems With Friction and Impacts.
LDR:03709nmm a2200397 4500 001 2400058
005 20240924101848.5
006 m o d
007 cr#unu||||||||
008 251215s2023 ||||||||||||||||| ||eng d
020 $a 9798380384766
035 $a (MiAaPQ)AAI30568375
035 $a AAI30568375
040 $a MiAaPQ $c MiAaPQ
100 1 $a Halm, Mathew. $3 3770030
245 1 0 $a Addressing Stiffness-Induced Challenges in Modeling and Identification for Rigid-Body Systems With Friction and Impacts.
260 1 $a Ann Arbor : $b ProQuest Dissertations & Theses, $c 2023
300 $a 174 p.
500 $a Source: Dissertations Abstracts International, Volume: 85-03, Section: B.
500 $a Advisor: Posa, Michael.
502 $a Thesis (Ph.D.)--University of Pennsylvania, 2023.
520 $a Imperfect, useful dynamical models have enabled significant progress in planning and controlling robotic locomotion and manipulation. Traditionally, these models have been physics-based, with accuracy relying upon manual calibration only feasible in laboratory environments. As robotics expands into complex real-world applications, models of unknown environments must instead be automatically fit to limited data. One major challenge is modeling frictional contact, especially during collisions involved in common robotics tasks. Rapid deformation under impact manifests as extreme sensitivity to initial conditions and material properties. Thus, even slight errors in state estimation and system identification can lead to significant prediction errors. Consequently, model inaccuracy or the sim-to-real gap often hinders the development of performant robotics algorithms.Physical models can be optimized using advanced techniques to overcome these challenges, but such methods have limited tractability when a large number of parameters are unknown. Furthermore, even given accurate parameters, roboticists often make inaccurate rigid-body approximations to reduce the computational burdens of physical simulation to meet faster-than-real-time requirements. An alternative black-box approach, in which models are learned from scratch, has attempted to address these issues for instance using deep neural networks (DNN's). While DNNs in theory can capture any dynamical behavior, they empirically struggle with the stiff behaviors associated with contact.This dissertation instead focuses on scaling physical model identification to the high-dimensional setting and quantifying the limited accuracy of low-fidelity models. We consider rigid bodies under-going rigid contact, for which infinite stiffness is represented as constrained optimization. By careful treatment of these constraints, we demonstrate that infinitely-stiff dynamics can be identified by optimizing a non-stiff objective. In conjunction, we use DNN's in a white-box setting to model physical quantities, specifically reconstructing geometries from scratch. We then consider how rigid-body collision models lack the fidelity to correctly predict outcomes of nearly-simultaneous impacts-such as heel and toe strikes during a footstep. We develop a theoretical basis to capture partial knowledge of impact events as uncertain set-valued outcomes, and again use numerical optimization to compute approximations of such sets.
590 $a School code: 0175.
650 4 $a Robotics. $3 519753
650 4 $a Mechanics. $3 525881
650 4 $a Computer science. $3 523869
650 4 $a Mechanical engineering. $3 649730
653 $a Robotics algorithms
653 $a Rigid-body systems
653 $a Deep neural networks
653 $a Robotic locomotion
690 $a 0771
690 $a 0346
690 $a 0800
690 $a 0984
690 $a 0548
710 2 $a University of Pennsylvania. $b Mechanical Engineering and Applied Mechanics. $3 3174099
773 0 $t Dissertations Abstracts International $g 85-03B.
790 $a 0175
791 $a Ph.D.
792 $a 2023
793 $a English
856 4 0 $u https://pqdd.sinica.edu.tw/twdaoapp/servlet/advanced?query=30568375