Parametric modeling of deformable linear objects for robotic outfitting and maintenance of space systems

Area of Science:

• Robotics and Automation
• Aerospace Engineering
• Applied Physics

Background:

• Long-duration space missions require autonomous robotic systems for construction and maintenance.
• Robotic manipulation planning has focused on rigid objects, neglecting Deformable Linear Objects (DLOs) like electrical cables.
• Existing DLO research often overlooks cables' non-zero bend equilibrium configurations.

Purpose of the Study:

• To develop a model-based optimization approach for estimating cable configurations in space.
• To address the challenge of robotic manipulation for flexible electrical cables in space architectures.
• To improve the accuracy and efficiency of modeling cable behavior for autonomous systems.

Main Methods:

• Implemented a model-based optimization approach with a discretization level parameter for accuracy-complexity trade-offs.
• Utilized observed 2D cable configurations for parameter estimation to refine the cable model.
• Validated the parameter estimation by comparing predicted configurations to real cable data.

Main Results:

• The model-based optimization approach effectively estimates cable configurations.
• Parameter estimation reduced prediction errors by an order of magnitude compared to previous methods.
• The method accounts for non-zero equilibrium configurations, a critical factor for real-world cables.

Conclusions:

• Accurate modeling of flexible cables is crucial for robotic outfitting and maintenance in space.
• The developed method enhances the state-space representation of cable payloads for robotic control.
• This research advances the capabilities of autonomous systems for complex in-space servicing operations.