Research on model-free adaptive trajectory tracking cooperative control method of dual-arm

Abstract

Addressing the problem of trajectory tracking accuracy degradation caused by high-degree-of-freedom dynamic coupling, model uncertainty, and communication delay in dual-arm cooperative systems, this paper presents a model-free adaptive control method that utilizes partial form dynamic linearization (PFDL-MFAC). A hierarchical control architecture combining kinematic decoupling with distributed cooperation decomposes the complex high-degree-of-freedom coupled system into parallel joint-space optimization processes, thereby reducing computational complexity. The multi-step historical input increment is introduced to compensate for the joint drive lag effect, while the online pseudo-gradient estimation with a reset mechanism is incorporated to improve the tracking robustness under time-varying delays. The proposed global unified parameter strategy verifies the algorithm’s adaptability to degree-of-freedom expansion and heterogeneous joints. The simulation results confirm its stable tracking performance under ideal conditions. Comparative experiments under random delay conditions demonstrate that the proposed control scheme enhances tracking precision over traditional PID control methods, reducing the steady-state error and peak displacement. Experimental evidence shows that the algorithm has strong robustness against time-delay disturbances. This method offers a novel solution for high-precision cooperative control of networked robotic systems operating under both time-delay constraints and dynamic uncertainties.