Wigand, Dennis Leroy: Modeling robot control systems in compliant interaction with the environment. Bridging the gap between the envisioned task and the robot’s behavior. 2022
Inhalt
- Abstract
- Acknowledgments
- Contents
- listofillustrations
- List of figures
- List of tables
- List of listings
- 1 Introduction
- 1.1 On Compliant Interaction
- 1.2 Problem Statement
- 1.3 Research Approach
- 1.4 Goal and Research Questions
- 1.5 Contributions and Outline
- 1.6 Publications
- 2 Engineering Compliant Control Systems
- 2.1 Task Description and Task-Level Programming
- 2.2 Compliant Interaction Control
- 2.2.1 Reacting to External Disturbances
- 2.2.2 Multi-Objective Control
- 2.2.3 Control in Contact Situations
- 2.3 Component-Based Software Engineering for Robotics
- 2.3.1 Separation of Concerns
- 2.3.1.1 Separation of Concerns: C1 Computation
- 2.3.1.2 Separation of Concerns: C2 Communication
- 2.3.1.3 Separation of Concerns: C3 Coordination
- 2.3.1.4 Separation of Concerns: C4 Configuration
- 2.3.1.5 Separation of Concerns: C5 Composition
- 2.3.2 Separation of Roles
- 2.4 Model-Driven Engineering for Robotics
- 2.5 Conclusion
- CoSiMA
- 3 Modeling Robot Control Systems
- 3.1 L3Dim: Layered 3 Dimensions Approach
- 3.1.1 Vertical Layers
- 3.1.2 Horizontal Dimensions
- 3.1.3 Guidelines to Support Language Evolution for LM&C
- 3.2 L3Dim's Modularization and Composition Mechanisms
- 3.2.1 Composing Horizontal Dimensions
- 3.2.1.1 Capability Dimension
- 3.2.1.2 Hardware Platform Dimension
- 3.2.1.3 Software Platform Dimension
- 3.2.2 Linking Vertical Layers
- 3.3 Compliant Simulation and Modeling Architecture (CoSiMA)
- 3.3.1 Platform-Independent Capabilities
- 3.3.2 Robots as Hardware Platform
- 3.3.3 OROCOS RTT as Software Platform
- 3.3.4 Modular Generator Composition for OROCOS RTT
- 3.4 Evaluation
- 3.4.1 Qualitative Evaluation of L3Dim and CoSiMA
- 3.4.1.1 Language Modules for Composition ([g:1]G1)
- 3.4.1.2 Exchangeability of Heterogeneous Modules ([g:2]G2)
- 3.4.1.3 Well-Defined Interfaces for LM&C ([g:3]G3)
- 3.4.1.4 Limiting the Impact of Language Evolution ([g:4]G4)
- 3.4.2 Quantitative Evaluation of CoSiMA
- 3.5 Conclusion
- 4 Timing Modeling in CoSiMA
- 4.1 Domain Analysis
- 4.2 Modeling of Timing Constraints
- 4.2.1 WCET Modeling
- 4.2.2 WCE2ERT and Sense-React Chain Modeling
- 4.2.3 Core Affinity and Schedule Modeling
- 4.3 Synthesis of an Executable Schedule
- 4.4 Evaluation
- 4.5 Conclusion
- Compliant Interaction
- 5 CI Domain Analysis and Modeling
- 5.1 Task Description for CI
- 5.2 Contact-Based Interactions
- 5.2.1 Compliant Interaction Modeling via Contacts
- 5.2.2 Expressing Contact Couplings via Constraints
- 5.2.3 Contact Surface Constraints
- 5.2.4 Kinematic Chains and Joint Constraints
- 5.2.5 Virtual Manipulator Constraints
- 5.3 Contact Situations and Transitions
- 5.4 Conclusion
- 6 CI Language Design and (Meta-)Modeling
- 6.1 Related Work
- 6.2 Modeling Contact-Based Compliant Interactions
- 6.2.1 Physical Environment Model
- 6.2.2 Modeling Contacts and Constraints
- 6.2.3 VM Constraint with Internal Wrench Handling
- 6.2.4 Modeling Constraint Prioritization Structures
- 6.3 CI Composition and other Task Aspects
- 6.4 Discussion
- 6.4.1 Integration into CoSiMA's Composition Structure
- 6.4.2 Comparison with Related Approaches
- 6.4.3 Decoupling Coordination and CI
- 6.5 Conclusion
- 7 CI Architecture Synthesis
- Perspectives
- 8 Evaluation
- 8.1 Tetra-Arm Object Handling
- 8.2 Dual-Arm Yoga Mat Rolling
- 8.3 Single-Arm Clamp Assembly
- 8.4 Conclusion on the Modeling of CI
- 8.5 Conclusion on the Execution of CI
- 9 Conclusion
- A Appendix
- A.1 Generation of Component-Level Behavior
- A.2 Generation of a Robot Interface Component Configuration
- A.3 Generation of a Scheduling Configuration
- A.4 CI World and Constraint Models
- A.5 Timing Constraints and Schedule
- Acronyms
- Glossary
- Bibliography
- Declaration
- Colophon