Detail View
Peg-in-Hole Insertion: A Hierarchical Control Framework with Environment-Adaptive Zero-Moment Strategy
WEB OF SCIENCE
SCOPUS
- Title
- Peg-in-Hole Insertion: A Hierarchical Control Framework with Environment-Adaptive Zero-Moment Strategy
- Alternative Title
- 펙인홀 삽입: 환경 적응형 영 모멘트 제어 전략을 적용한 계층 제어
- DGIST Authors
- Jeong Juwon ; Oh Sehoon ; Hwang Minho
- Advisor
- 오세훈
- Co-Advisor(s)
- Hwang Minho
- Issued Date
- 2025
- Awarded Date
- 2025-08-01
- Type
- Thesis
- Description
- Peg-in-hole, Contact-rich manipulation task, multi-task control, Hierarchical control framework, Disturbance observer, Force/moment feedback control
- Abstract
-
The Peg-in-Hole insertion task is a fundamental operation with wide-ranging applications in industrial and everyday scenarios. Because the insertion environment is difficult to model precisely, position- and orientation errors between the peg and hole are inevitable, and achieving stable force/moment control in the contact-rich environment is particularly challenging.
더보기
This paper proposes a task-space manipulator control strategy that enables reliable peg-in-hole insertion under tight geometric tolerances. I demonstrate that the coupling among the subtasks’ controllers has a critical impact on task performance through case studies on the task hierarchization and the priority setting. Building on these findings, a variable center-of-rotation scheme to address changing environmental constraints and a zero-moment control method exploiting a 6-axis force/torque sensor. And a disturbance-observer-based control architecture that explicitly incorporates task-space inertia terms to maintain robust force and moment regulation in high-precision environments. In addition, a disturbance-observer-based framework incorporating task-space inertial terms enables stable force/moment control even under tight tolerance requirements. The effectiveness of the proposed framework is demonstrated in both high-fidelity simulations and real-robot experiments.
The study provides new design guidelines for PiH-oriented controllers and is expected to generate synergy when integrated with learning-based or multi-sensor high-level control strategies.|펙인홀 삽입 작업은 산업 전반뿐만 아니라 실생활에서도 다양하게 활용될 수 있는 핵심 기술이다. 하지만 삽입 환경을 정확히 모델링하기 어렵기 때문에 펙과 홀 사이의 위치 및 자세 오차 발생이 불가피하며, 다중 접촉 환경에 안정적으로 힘/모멘트 제어 구현 또한 까다롭다.
본 논문은 해당 작업을 성공적으로 수행하기 위해 작업 공간 기반의 매니퓰레이터 제어 전략을 제안한다. 여러 하위 제어기 간의 간섭이 작업 성능에 미치는 영향을 가정하고, 계층 제어 적용 여부 및 제어기 우선 순위 설정에 대한 사례 연구를 통해 해당 가정을 실험적으로 검증했다. 또한 Peg-in-Hole 삽입 작업의 특성을 반영하여, 환경 변화에 적응 가능한 가변 회전 중심 방법론과 6축 힘/토크 센서를 이용한 새로운 모멘트 피드백 제어 전략을 제시한다. 이 때 공차 정밀도가 높은 상황에서도 안정적인 힘/모멘트 제어가 가능하도록 작업 공간의 관성 항이 고려된 새로운 외란 관측기 기반의 제어 구조를 제시한다. 제안된 제어기의 유효성은 시뮬레이션과 실험을 통해 검증되었다.
본 연구는 펙인홀 삽입 작업 수행을 위한 새로운 제어기 설계 가이드라인을 제시하고 있다. 향후 학습 기반 또는 다중 센서 융합형 상위 제어 방법론과 접목될 경우, 시너지 효과가 기대된다.
- Table Of Contents
-
1 Introduction 1
1.1 Problem Statement 1
1.2 Related Works 2
1.3 Thesis Overview 3
2 Zero-Moment Control in Tight Contact-Rich Environments 4
2.1 Zero-Moment Control for Resolving Jamming 4
2.2 Disturbance-Observer-Based Control Framework for Zero-Moment Control 5
2.2.1 Simple Closed-loop Feedback Control 5
2.2.2 Disturbance-Observer-based Zero-Moment Control 6
2.2.3 Dynamic Disturbance-Observer-based Zero-Moment Control 7
2.3 Experimental Validation 7
3 Environment-Dependent Variability of the Peg’s Center of Rotation 9
3.1 Center of Rotation 9
3.2 Variable Center of Rotation at the Peg Tip 10
4 Controller Design within a Hierarchical Control Framework 12
4.1 Subtasks for Peg-in-Hole Insertion 12
4.2 Null-Space Projection in a Hierarchical Control Framework 13
4.2.1 Multi-Task Control without considering Task Coupling 13
4.2.2 Multi-Task Control considering Task Coupling 14
4.3 Subtask Priority Setting 14
5 Experimental Verification on Peg-in-Hole Insertion 15
5.1 Environmental Setup 15
5.2 Task-space Control Scheme for Manipulators 16
5.2.1 Translation-Z Dimension: Feedforward Force Control 16
5.2.2 Translation-X/Y Dimensions: Position Feedback Control 16
5.2.3 Rotation Dimensions: Zero-Moment Control 17
5.3 Experimental Results 17
5.3.1 Center of Rotation 17
5.3.2 Priority Setting 19
6 Conclusion 23
- URI
-
https://scholar.dgist.ac.kr/handle/20.500.11750/59846
http://dgist.dcollection.net/common/orgView/200000893125
- Degree
- Master
- Department
- Artificial Intelligence Major
- Publisher
- DGIST
File Downloads
- There are no files associated with this item.
공유
Total Views & Downloads
???jsp.display-item.statistics.view???: , ???jsp.display-item.statistics.download???:
