Leg-adjustment strategies for stable running in three dimensions

Area of Science:

• Biomechanics
• Robotics
• Locomotion Analysis

Background:

• The spring-loaded inverted pendulum (SLIP) model effectively describes 2D human and animal running dynamics.
• SLIP models demonstrate inherent stability and perturbation recovery without active force control, relying on leg angle adjustments.
• Extending running models to three dimensions (3D) is crucial for understanding complex locomotion.

Purpose of the Study:

• To investigate novel feed-forward leg adjustment strategies for 3D running using an extended SLIP model.
• To explore the influence of different reference axes, including the center of mass (CoM) velocity vector, on running stability.
• To assess the potential of these strategies for robotic locomotion and comparison with experimental data.

Main Methods:

• Developed a 3D SLIP (3D SLIP) model incorporating six leg-adjustment strategies.
• Introduced the CoM velocity vector as a movement-related reference axis alongside fixed global axes.
• Analyzed model stability, parameter domains, and directional stability under various leg-adjustment strategies.

Main Results:

• Leg-adjustment strategies incorporating the CoM velocity vector yielded stable 3D running with large stability domains.
• Velocity-based strategies demonstrated asymptotic and neutral directional stability, crucial for perturbation recovery.
• The model predicted stable running at arbitrary low speeds and smooth transitions from 3D to 2D/1D SLIP dynamics.
• A specific velocity-based strategy showed a large parameter region for robust hopping and running, suitable for engineering applications.

Conclusions:

• Movement-related leg-adjustment strategies, particularly those using the CoM velocity vector, are key to stable 3D running.
• These findings suggest that compliant leg function and internal movement references enable 3D running in both biological and artificial systems.
• The 3D SLIP model with velocity-based control offers a promising framework for understanding and implementing robust robotic locomotion.