Mode-selective adaptive admittance force-position control for safe robotic interaction in unknown time-varying...

Area of Science:

• Robotics
• Control Systems Engineering
• Artificial Intelligence

Background:

• Robotic manipulators face challenges in complex, dynamic environments, requiring safe interaction and precise trajectory tracking.
• Tasks involve transitions between free and contact motion, with uncertain parameters and impact-induced oscillations.

Purpose of the Study:

• To propose a mode-selective adaptive admittance force-position control framework for robotic manipulators.
• To address challenges in safe interaction, trajectory tracking, and force control in unknown, time-varying environments.

Main Methods:

• An activation matrix for mode selection between force and position control subspaces.
• An environment-aware adaptive admittance outer loop using Gaussian Process Regression-enhanced Extended Kalman Filter (GPR-EKF) and Lyapunov-based adaptation.
• A mode-gated self-tuning PID inner loop for enhanced trajectory tracking.

Main Results:

• Reduced force overshoot during robotic interactions.
• Improved steady-state force tracking accuracy.
• Enhanced safety and performance in unknown, time-varying environments.

Conclusions:

• The proposed control framework effectively manages complex robotic interaction tasks.
• The method demonstrates significant improvements in force control and trajectory tracking accuracy.
• This approach advances safe and reliable human-robot collaboration in dynamic settings.