Stable Walking of a Biped Robot Controlled by Central Pattern Generator Using Multivariate Linear Mapping.

Area of Science:

• Robotics
• Control Systems
• Biomechanics

Background:

• Bipedal robots require robust walking controllers for stability in varied environments.
• Existing Central Pattern Generator (CPG) models can be complex, hindering widespread application.
• Improving CPG model efficiency and stability is crucial for advanced robotic locomotion.

Purpose of the Study:

• To propose a new, simplified Central Pattern Generator (CPG) walking controller for biped robots.
• To enhance the stability and versatility of bipedal walking across different terrains.
• To reduce the computational complexity of CPG models for practical implementation.

Main Methods:

• Designed a lower limb mechanical structure and established a hybrid dynamic model using Lagrange and angular momentum conservation.
• Determined passive walking initial values via Poincaré mapping and cell mapping.
• Developed a lightweight CPG model using multivariate linear mapping based on a Hopf oscillator.

Main Results:

• A new CPG model was established with a preliminary parameter-tuning strategy.
• Joint simulations in MATLAB and V-REP demonstrated stable periodic gaits for the biped robot.
• The controller proved effective in both flat and uphill walking scenarios.

Conclusions:

• The proposed multivariate linear mapping-based CPG controller significantly improves biped robot walking stability and versatility.
• This simplified CPG approach offers a valuable reference for CPG generation and tuning in robotics research.
• The method provides a pathway for more adaptable and efficient bipedal locomotion systems.