A Hybrid Force-Position Strategy for Shape Control of Deformable Linear Objects With Graph Attention Networks

2025-08-10

AI Summary

Controlling deformable linear objects (DLOs) remains challenging due to their high degrees of freedom, strong nonlinearity, and underactuation. To address this, we propose a force–position hybrid model predictive control (MPC) framework. Our method integrates force-space trajectory planning with position-space MPC and introduces a graph attention network (GAT) to explicitly encode actions and learn the underlying dynamics—enabling geometric structure awareness and adaptive extraction of object-specific attributes. This design significantly improves modeling accuracy and control robustness. We validate the approach in both simulation and real-world robotic experiments, demonstrating precise and stable dynamic shape regulation of DLOs. All code and experimental videos are publicly released, establishing a reproducible, generalizable paradigm for DLO manipulation.

Technology Category

• Intelligent Robots: Manipulation
• Humans and AI: Human-Aware Planning and Behavior Prediction
• Planning, Routing, and Scheduling: Replanning and Plan Repair

Application Category

• Graph Algorithms and Modeling for the Web: Efficient manipulation of static and dynamic Web-related graphs
• Systems and Infrastructure for Web, Mobile and WoT: Applied ML and AI for Web-based mobile applications
• Responsible Web: Machine-in-the-loop, human agency and autonomy

Abstract

Manipulating deformable linear objects (DLOs) such as wires and cables is crucial in various applications like electronics assembly and medical surgeries. However, it faces challenges due to DLOs' infinite degrees of freedom, complex nonlinear dynamics, and the underactuated nature of the system. To address these issues, this paper proposes a hybrid force-position strategy for DLO shape control. The framework, combining both force and position representations of DLO, integrates state trajectory planning in the force space and Model Predictive Control (MPC) in the position space. We present a dynamics model with an explicit action encoder, a property extractor and a graph processor based on Graph Attention Networks. The model is used in the MPC to enhance prediction accuracy. Results from both simulations and real-world experiments demonstrate the effectiveness of our approach in achieving efficient and stable shape control of DLOs. Codes and videos are available at https://sites.google.com/view/dlom.

Problem

• Control deformable linear objects' complex nonlinear dynamics
• Address infinite degrees of freedom in DLO manipulation
• Overcome underactuated system challenges in shape control

Innovation

• Hybrid force-position strategy for DLO shape control
• Graph Attention Networks for dynamics modeling
• Model Predictive Control enhances prediction accuracy