Combined Feedback Feedforward Control of a 3-Link Musculoskeletal System Based on the Iterative Training Method

Area of Science:

• Robotics
• Biomechanics
• Control Systems

Background:

• Human arm movement inspires humanoid robot design.
• Musculoskeletal systems exhibit joint redundancy, posing control challenges.
• Redundancy leads to non-unique joint configurations for desired end-effector paths.

Purpose of the Study:

• To apply an iterative learning controller to a 3-link musculoskeletal system with 6 muscles.
• To address joint redundancy and control challenges in robotic arms.
• To improve trajectory tracking accuracy in noisy environments.

Main Methods:

• An iterative learning controller was designed for a 3-link musculoskeletal robot.
• The controller utilized task space as feedforward and muscle space as feedback.
• A forgetting factor was implemented to manage noise and ensure convergence.

Main Results:

• Controller performance improved iteratively with learning steps.
• System errors decreased significantly over iterations.
• The end-effector trajectory closely matched the desired path after 1000 iterations.

Conclusions:

• Iterative learning control effectively manages joint redundancy in musculoskeletal robots.
• The proposed controller enhances trajectory tracking accuracy despite system noise.
• This approach offers a viable solution for controlling complex robotic systems inspired by human anatomy.