Adaptive Kinematic Control of Robot Manipulators: A Concurrent Learning Based Approach
Bildiriler > Tam Metin Bildiri
Yazarlar: Armin Razmgiri, Serhat Obuz, Enver TATLICIOĞLU, Erkan Zergeroglu, Erman SELİM
2025 11th International Conference on Control, Decision and Information Technologies (CoDIT), , , Uluslararası, Tam Metin Bildiri, ss 1043-1048
DOI: 10.1109/codit66093.2025.11321405
Özet
This paper presents an adaptive kinematic control strategy for robotic manipulators, which takes advantage of a concurrent learning-based approach to address kinematic uncertainties. Departing from typical approaches that rely on position-level inverse kinematics, the developed framework operates directly in Cartesian space, thereby reducing computational complexity and mitigating singularity-related issues. The control framework incorporates a concurrent learning-based adaptive update law, providing precise end-effector trajectory tracking and real-time identification of uncertain kinematic parameters under interval excitation condition, which is less stringent than persistent excitation. Stability analysis is conducted using a Lyapunov-Based framework which proves the global exponential convergence of both tracking and parameter estimation errors. Numerical simulations validate the effectiveness of the developed approach, accurately demonstrating trajectory tracking and identification of the uncertain kinematic terms.