Integral sliding mode control for an anthropomorphic finger based on nonlinear extended state observer

Area of Science:

• Robotics
• Control Systems Engineering
• Biomechanical Engineering

Background:

• Anthropomorphic fingers often struggle with stability and accuracy in joint motion control due to coupling disturbances.
• This limitation impacts the performance of complex grasping and manipulation tasks for robotic hands.

Purpose of the Study:

• To propose a novel control strategy for anthropomorphic fingers driven by pneumatic artificial muscles (PAMs).
• The goal is to achieve stable and accurate joint motion control, overcoming coupling disturbances.
• To enhance the steady-state performance of the finger's joint motion control.

Main Methods:

• A nonlinear extended state observer (NESO) was developed to estimate and compensate for coupling disturbances.
• An integral sliding mode controller (ISMC) was designed for precise joint motion control.
• Lyapunov methods were employed to mathematically prove the convergence and stability of the NESO and ISMC.

Main Results:

• Experimental validation demonstrated the effectiveness of the proposed control strategy.
• The NESO successfully observed and compensated for coupling disturbances in the anthropomorphic finger.
• The ISMC achieved stable and accurate joint motion control, improving steady-state performance.

Conclusions:

• The proposed control strategy, combining NESO and ISMC, significantly improves the stability and accuracy of anthropomorphic finger joint motion control.
• This approach effectively addresses the challenges posed by coupling disturbances in PAM-driven systems.
• The findings are crucial for advancing the capabilities of anthropomorphic hands in complex manipulation tasks.