• Proposes a novel RBFNNA-HSMC method that integrates radial basis function neural networks with hierarchical sliding mode control to address trajectory tracking of elastic tendon-driven manipulators under model uncertainty and disturbances.
• Demonstrates superior tracking accuracy compared to traditional model-based HSM control, with experimental maximum tracking errors below 2.593×10-3 rad and 1.624×10-3 rad for double-joint trajectory tracking.
• Establishes closed-loop stability via Lyapunov stability theory, ensuring robustness against model inaccuracies and external disturbances.
• Validates the effectiveness and adaptability of the proposed control method through simulations and experiments on a two-DOF ETDM, highlighting its potential for applications in human-robot collaboration and flexible robotics.