東華大學圖書館 |

語系: 繁體中文

說明(常見問題)

回圖書館首頁

手機版館藏查詢

登入

回首頁 到查詢結果 [ null ]

切換: 標籤 | MARC模式 | ISBD

Towards Generalizable Contact-Rich M...

Zhang, Xiang,

FindBook

Google Book

Amazon

博客來

Towards Generalizable Contact-Rich Manipulation With Robust Skill Learning, Safe Compliance Tuning, and Efficient Online Adaptation /

紀錄類型:	書目-電子資源 : Monograph/item
正題名/作者:	Towards Generalizable Contact-Rich Manipulation With Robust Skill Learning, Safe Compliance Tuning, and Efficient Online Adaptation // Xiang Zhang.
作者:	Zhang, Xiang,
面頁冊數:	1 electronic resource (140 pages)
附註:	Source: Dissertations Abstracts International, Volume: 86-09, Section: B.
Contained By:	Dissertations Abstracts International86-09B.
標題:	Robotics. -
電子資源:	https://pqdd.sinica.edu.tw/twdaoapp/servlet/advanced?query=31635452
ISBN:	9798310103290

Towards Generalizable Contact-Rich Manipulation With Robust Skill Learning, Safe Compliance Tuning, and Efficient Online Adaptation /
Zhang, Xiang,

Towards Generalizable Contact-Rich Manipulation With Robust Skill Learning, Safe Compliance Tuning, and Efficient Online Adaptation /Xiang Zhang. - 1 electronic resource (140 pages)

Source: Dissertations Abstracts International, Volume: 86-09, Section: B.

Robotic manipulation is increasingly transitioning from simple trajectory-following tasks to more complex, contact-rich applications such as assembly, pivoting, polishing, and wire harnessing. These tasks require robots to actively interact with objects by applying appropriate forces, posing significant challenges for automation due to the complexities of contact dynamics, safety concerns, and the need for generalization across different objects and environments. To address these challenges, this dissertation focuses on developing efficient, adaptive methodologies for enhancing robots' real-time interaction capabilities, particularly in contact-rich and force-sensitive tasks.The central theme of this dissertation is centered around three key challenges:Learning Robust Manipulation Motions for Contact-Rich Manipulation: A major challenge in robotics is to train robots to efficiently manipulate objects while making and controlling contact. Part I of the dissertation introduces techniques for improving the training process of contact-rich manipulation by incorporating human prior knowledge and optimizing skill acquisition in both simulated and real-world environments. These methods aim to reduce the sim-to-real gap, ensuring that robots can perform manipulation tasks robustly and with greater data efficiency.Compliance Gain Tuning for Safe Robot-Environment Interaction: Robots must interact safely and effectively with their environment, especially during contact-rich tasks. Part II of the dissertation addresses this by presenting algorithms for real-time compliance tuning, allowing robots to dynamically adjust their stiffness based on sensory feedback. These methods ensure that robots can safely apply the necessary forces during manipulation without compromising the safety of the task or the environment.Online Skill Adaptation for Generalizable Robot Skills: As task settings and environmental conditions change, robots must be able to generalize their learned skills to new scenarios. Part III of the dissertation combines offline skill learning with real-time online adaptation, allowing robots to adjust their manipulation skills based on live sensor data. This section emphasizes the development of policies that enable robots to perform contact-rich tasks, such as assembly and insertion, across different object types and dynamic environments.Through addressing these three core challenges, this dissertation contributes to the advancement of robust and adaptive robot manipulation techniques, ultimately pushing the boundaries of what robots can achieve in real-world settings.

English

ISBN: 9798310103290Subjects--Topical Terms:

519753
Robotics.
Subjects--Index Terms:

Autonomous systems

Towards Generalizable Contact-Rich Manipulation With Robust Skill Learning, Safe Compliance Tuning, and Efficient Online Adaptation /
LDR:04129nmm a22004333i 4500 001 2432257
005 20250707081329.5
006 m o d
007 cr|nu||||||||
008 261011s2024 miu||||||m |||||||eng d
020 $a 9798310103290
035 $a (MiAaPQD)AAI31635452
035 $a AAI31635452
040 $a MiAaPQD $b eng $c MiAaPQD $e rda
100 1 $a Zhang, Xiang, $e author. $0 (orcid)0009-0006-0496-4458 $3 3817727
245 1 0 $a Towards Generalizable Contact-Rich Manipulation With Robust Skill Learning, Safe Compliance Tuning, and Efficient Online Adaptation / $c Xiang Zhang.
264 1 $a Ann Arbor : $b ProQuest Dissertations & Theses, $c 2024
300 $a 1 electronic resource (140 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: 86-09, Section: B.
500 $a Advisors: Tomizuka, Masayoshi Committee members: Horowitz, Roberto; Abbeel, Pieter.
502 $b Ph.D.Eng. $c University of California, Berkeley $d 2024.
520 $a Robotic manipulation is increasingly transitioning from simple trajectory-following tasks to more complex, contact-rich applications such as assembly, pivoting, polishing, and wire harnessing. These tasks require robots to actively interact with objects by applying appropriate forces, posing significant challenges for automation due to the complexities of contact dynamics, safety concerns, and the need for generalization across different objects and environments. To address these challenges, this dissertation focuses on developing efficient, adaptive methodologies for enhancing robots' real-time interaction capabilities, particularly in contact-rich and force-sensitive tasks.The central theme of this dissertation is centered around three key challenges:Learning Robust Manipulation Motions for Contact-Rich Manipulation: A major challenge in robotics is to train robots to efficiently manipulate objects while making and controlling contact. Part I of the dissertation introduces techniques for improving the training process of contact-rich manipulation by incorporating human prior knowledge and optimizing skill acquisition in both simulated and real-world environments. These methods aim to reduce the sim-to-real gap, ensuring that robots can perform manipulation tasks robustly and with greater data efficiency.Compliance Gain Tuning for Safe Robot-Environment Interaction: Robots must interact safely and effectively with their environment, especially during contact-rich tasks. Part II of the dissertation addresses this by presenting algorithms for real-time compliance tuning, allowing robots to dynamically adjust their stiffness based on sensory feedback. These methods ensure that robots can safely apply the necessary forces during manipulation without compromising the safety of the task or the environment.Online Skill Adaptation for Generalizable Robot Skills: As task settings and environmental conditions change, robots must be able to generalize their learned skills to new scenarios. Part III of the dissertation combines offline skill learning with real-time online adaptation, allowing robots to adjust their manipulation skills based on live sensor data. This section emphasizes the development of policies that enable robots to perform contact-rich tasks, such as assembly and insertion, across different object types and dynamic environments.Through addressing these three core challenges, this dissertation contributes to the advancement of robust and adaptive robot manipulation techniques, ultimately pushing the boundaries of what robots can achieve in real-world settings.
546 $a English
590 $a School code: 0028
650 4 $a Robotics. $3 519753
650 4 $a Mechanical engineering. $3 649730
650 4 $a Automotive engineering. $3 2181195
653 $a Autonomous systems
653 $a Compliance control
653 $a Robot manipulation
653 $a Robotic systems
653 $a Online adaptation
690 $a 0771
690 $a 0548
690 $a 0540
710 2 $a University of California, Berkeley. $b Mechanical Engineering. $e degree granting institution. $3 3821326
720 1 $a Tomizuka, Masayoshi $e degree supervisor.
773 0 $t Dissertations Abstracts International $g 86-09B.
790 $a 0028
791 $a Ph.D.Eng.
792 $a 2024
856 4 0 $u https://pqdd.sinica.edu.tw/twdaoapp/servlet/advanced?query=31635452