High-precision dynamic torque control of high stiffness actuator for humanoids

Area of Science:

• Robotics
• Control Systems Engineering
• Mechatronics

Background:

• High stiffness actuators (HSAs) are crucial for the motion performance of torque-controlled humanoid robots.
• Precise torque control is essential for achieving high performance in robotic systems with HSAs.
• Existing control methods may struggle with nonlinearities and noisy signals in HSA torque control.

Purpose of the Study:

• To propose a high-precision dynamic torque control (HDTC) method for HSAs in humanoid robots.
• To enhance the motion performance and control accuracy of torque-controlled humanoid robots.
• To address challenges posed by nonlinear coupling and noisy sensor data in HSA control.

Main Methods:

• Developed a novel dynamic current control to linearize the HSA torque control system by estimating and compensating nonlinear coupling.
• Designed an enhanced internal model control to ensure high tracking accuracy, robust to noisy torque signals and numerical differentiation.
• Integrated dynamic current control and enhanced internal model control into the proposed HDTC framework.

Main Results:

• The proposed HDTC method demonstrates accurate and adaptable torque control for HSAs.
• Comparative experiments verified the superiority of the HDTC over existing methods.
• The control system effectively handles nonlinearities and noisy signals, improving overall robot motion performance.

Conclusions:

• The proposed HDTC offers a significant advancement in controlling HSAs for humanoid robots.
• The method enhances robotic system accuracy and adaptability, paving the way for more sophisticated humanoid robot applications.
• HDTC provides a robust solution for high-performance torque control in challenging robotic environments.