Design and Control of a Bio-inspired Wheeled Bipedal Robot

Overview

• This paper focuses on improving the dynamic performance of wheeled bipedal robots (WBRs) through innovations in both hardware and software.
• The researchers propose and implement a bionic mechanical design inspired by the human barbell squat, which distributes the weight onto the robot's hip and knee joints.
• They also devise a novel model-based controller that combines a height-variable wheeled linear inverted pendulum (HV-wLIP) model, Control Lyapunov Function (CLF), and whole-body dynamics for theoretically guaranteed stability and efficient computation.

Plain English Explanation

Wheeled bipedal robots are machines that can move around on two wheels like humans walk on two legs. This paper aims to make these robots move in a more agile and versatile way. The researchers were inspired by the human barbell squat exercise and designed a new mechanical structure for the robot that distributes its weight onto the hip and knee joints. This helps the robot's motors be more effective while still allowing the robot to move its base around a lot.

The researchers also developed a new control system for the robot that uses a special mathematical model called the height-variable wheeled linear inverted pendulum (HV-wLIP) model, as well as other techniques, to make the robot's movements more stable and efficient. Experiments showed that this robot can do deep squats like a human and maintain control of its center of mass even when dealing with external disturbances or unknown terrain.

Technical Explanation

The proposed bionic mechanical design distributes the robot's weight onto its hip and knee joints, improving the effectiveness of the joint motors while maintaining a relatively large workspace for the base link (see Fig. 1). This is inspired by the human barbell squat exercise.

The researchers also developed a novel model-based controller that synthesizes the HV-wLIP model, Control Lyapunov Function (CLF), and whole-body dynamics. This provides theoretically guaranteed stability and efficient computation compared to other control approaches. The HV-wLIP model is a more accurate approximation of WBR dynamics, enabling more agile response and providing a solid theoretical basis for the controller design.

Experimental results demonstrate the robot's capabilities, including performing human-like deep squats and maintaining tracking of the center of mass (CoM) velocity while manipulating the base states. The robot also exhibited robustness against external disturbances and unknown terrains.

Critical Analysis

The paper provides a comprehensive approach to improving the dynamic performance of wheeled bipedal robots through both hardware and software innovations. The proposed bionic mechanical design and model-based control system appear to be effective, as demonstrated by the experimental results.

However, the paper does not extensively discuss the limitations of the approach or potential areas for further research. For example, it is unclear how the robot's performance would scale to larger or more complex tasks, or how the control system would handle more extreme environmental conditions or unforeseen disturbances.

Additionally, the paper does not compare the proposed system to other state-of-the-art approaches in the field of bipedal locomotion control or soft robotic control. Further analysis of the relative strengths and weaknesses of this approach compared to alternatives would help readers better understand its potential impact and applications.

Conclusion

This paper presents a novel approach to improving the dynamic performance of wheeled bipedal robots through advancements in both hardware and software. The bionic mechanical design and model-based control system demonstrated promising results in terms of agility, stability, and robustness. While the paper does not extensively explore the limitations of the approach, the proposed innovations have the potential to significantly advance the field of bipedal robotics and enable more versatile and capable robotic systems for a variety of applications.