Muscle Coordination Control for an Asymmetrically Antagonistic-Driven Musculoskeletal Robot Using Attractor Selection

Area of Science:

• Robotics
• Biomechanical Engineering
• Control Systems

Background:

• Musculoskeletal robots aim to mimic human/animal flexibility and dexterity.
• Complex actuator arrangements pose significant challenges for robot control system design.
• Biologically inspired control methods are crucial for advancing musculoskeletal robot development.

Purpose of the Study:

• To propose a muscle coordination control method using attractor selection for an antagonistic-driven musculoskeletal robot.
• To address challenges in controlling robots with asymmetric muscle arrangements.
• To evaluate the effectiveness of biologically inspired control for robotic systems.

Main Methods:

• Development of muscle coordination control models using virtual antagonistic muscle structures with symmetric arrangements.
• Application of attractor selection, a biologically inspired search method, to the control models.
• Comparison of control performance between the proposed method and a model without muscle coordination.

Main Results:

• The proposed muscle coordination control method demonstrated effectiveness in position control experiments.
• The virtual antagonistic muscle structure contributed to improved control performance.
• Attractor selection facilitated enhanced muscle coordination in the musculoskeletal robot.

Conclusions:

• The study validates the efficacy of attractor selection for muscle coordination in musculoskeletal robots.
• Biologically inspired control offers a promising approach for managing complex robotic systems.
• The proposed method enhances the dexterity and control of robots with asymmetric designs.