Human-Inspired Online Path Planning and Biped Walking Realization in Unknown Environment

Area of Science:

• Robotics
• Bio-inspired engineering
• Artificial intelligence

Background:

• Humanoid robot locomotion faces challenges in complex, unknown environments, limiting efficiency compared to humans.
• Current biped robots are sensitive to external changes and struggle with adaptive walking.
• Biological solutions offer promising approaches for efficient bipedal locomotion in challenging terrains.

Purpose of the Study:

• To present a human-inspired methodology for path planning and realization of bipedal walk in complex, unfamiliar environments.
• To develop a framework enabling humanoid robots to adapt their walking for obstacle avoidance and smooth path transitions.
• To bridge high-level path planning with low-level joint motion control for dynamic balance preservation.

Main Methods:

• Bio-inspired online path planning using clothoid curves for human-like paths.
• Automatic calculation of high-level gait parameters (step length, speed, direction, foot height) based on planned paths.
• Adaptation of motion primitives for low-level joint movement control and dynamic balance.

Main Results:

• A complete framework integrating path planning, gait parameter calculation, and motion primitive adaptation.
• Demonstrated ability for robots to adopt walking for obstacle avoidance and smooth transitions between paths.
• Enabled on-line modification of planned paths and walk parameters for real-time adaptation.

Conclusions:

• The proposed human-inspired framework significantly improves biped robot adaptability and efficiency in complex environments.
• The integration of path planning and motion primitives allows for dynamic balance and human-like locomotion.
• This methodology provides a robust solution for real-world humanoid robot applications requiring navigation in unfamiliar terrains.