Leg Locomotion Adaption for Quadruped Robots with Ground Compliance Estimation.

Area of Science:

• Robotics
• Control Systems
• Mechatronics

Background:

• Locomotion control for quadruped robots typically assumes rigid terrain.
• Real-world applications require robots to traverse diverse and compliant terrains.
• Existing control methods often lack adaptability to varying ground properties.

Purpose of the Study:

• To develop and evaluate an adaptive impedance control system for quadruped robot locomotion on compliant terrains.
• To enable robust foot-end trajectory control despite variations in ground compliance.
• To compare the performance of impedance control against pure position control in dynamic environments.

Main Methods:

• A single-leg robot prototype and test platform were constructed.
• Trajectory planning generated desired Cartesian foot-end coordinates.
• Geometric transformation converted Cartesian to virtual polar coordinates for impedance control.
• A ground compliance identification system estimated expected ground forces.
• An impedance controller adjusted control parameters based on identified compliance and trajectory deviations.

Main Results:

• The system successfully identified ground compliance in real-time.
• Accurate foot-end trajectory control was achieved across different ground compliances.
• Impedance control demonstrated superior performance compared to pure position control in adaptive locomotion.
• The proposed method enhanced robot stability and robustness on non-rigid surfaces.

Conclusions:

• Adaptive impedance control is crucial for robust quadruped robot locomotion on varied terrains.
• The developed ground compliance identification and impedance control system effectively enhances adaptability.
• This approach offers a promising solution for real-world robotic applications requiring terrain negotiation.