A Sensor Fusion Method for Pose Estimation of C-Legged Robots.

Area of Science:

• Robotics
• Mechatronics
• Control Systems

Background:

• Legged robots, particularly those with compliant legs, offer advantages over wheeled or tracked robots for navigating challenging environments like stairs and debris.
• Despite their locomotion capabilities,
• C
• legged robots lack developed algorithms for autonomous navigation.
• Accurate odometry estimation is crucial for implementing advanced navigation algorithms such as the Extended Kalman Filter.

Purpose of the Study:

• To present a novel algorithm for estimating the odometry of
• C
• legged robots with compliant legs.
• To analyze the robot's pose estimation using the proposed odometry algorithm.
• To enhance the autonomous navigation capabilities of
• C
• legged robots.

Main Methods:

• A new algorithm is proposed that utilizes leg encoders for improved robot localization.
• The algorithm employs a linear approximation of leg compression, reducing computational complexity compared to finite element analysis.
• Sensor fusion techniques are integrated with the odometry estimation for enhanced performance.

Main Results:

• The novel odometry estimation algorithm was successfully tested in both simulations and real-world robot experiments.
• The results demonstrate promising performance in estimating the robot's pose and odometry.
• The developed algorithm contributes to the advancement of autonomous navigation for
• C
• legged robots.

Conclusions:

• The proposed algorithm effectively estimates odometry for
• C
• legged robots with compliant legs.
• The method simplifies calculations by using a linear approximation, making it computationally efficient.
• The algorithm, combined with sensor fusion, provides a viable solution for enabling autonomous navigation in these robots.