Analytic Model for Quadruped Locomotion Task-Space Planning.

Area of Science:

• Robotics
• Biomechanics
• Computational Science

Background:

• Legged locomotion in animals is complex and challenging to replicate in robots.
• Dynamic locomotion planning leverages gravity but faces high computational costs with increased kinematic complexity.
• Real-time deployment in unstructured environments is hindered by computational demands.

Purpose of the Study:

• To propose a simplified method for generating center of mass and foot trajectories for quadruped robots.
• To address the computational challenges in real-time locomotion planning for legged robots.
• To enable complex robotic actions in unstructured environments.

Main Methods:

• A simplified model representing a quadruped as two connected bipeds.
• Extension of an algebraic bipedal locomotion model utilizing gravitational attractor topology.
• Generation of desired center of mass and foot trajectories.

Main Results:

• The developed model successfully generates trajectories consistent with previous research findings.
• The simplified approach shows promise in reducing computational load for trajectory optimization.
• The generated trajectories align with established bipedal locomotion strategies.

Conclusions:

• The proposed simplified quadruped locomotion model offers a viable approach to reduce computational complexity.
• This method serves as a foundational solution for whole-body trajectory optimization in robotics.
• Future deployment aims to achieve real-time planning for complex robotic actions in challenging terrains.