Introduction

We establish a wheeled-bipedal jumping dynamical (W-JBD) model to optimize the height control, and the Bayesian optimization for torque planning (BOTP) method based on a joint optimization framework for torque planning to achieve accurate height control and minimal energy cost.

ARM2021_Video.mp4

Citation

Y. Zhuang et al., "Height Control and Optimal Torque Planning for Jumping With Wheeled-Bipedal Robots," 2021 6th IEEE International Conference on Advanced Robotics and Mechatronics (ICARM), 2021, pp. 477-482, doi: 10.1109/ICARM52023.2021.9536196.

@INPROCEEDINGS{9536196,  
	author={Zhuang, Yulun and Xu, Yuan and Huang, Binxin and Chao, Mandan and Shi, Guowei and Yang, Xin and Zhang, Kuangen and Fu, Chenglong},  
	booktitle={2021 6th IEEE International Conference on Advanced Robotics and Mechatronics (ICARM)},   
	title={Height Control and Optimal Torque Planning for Jumping With Wheeled-Bipedal Robots},   
	year={2021},   
	pages={477-482},  
	doi={10.1109/ICARM52023.2021.9536196}}

Dependency

• Python 3.7
• Webots R2021a
• Advisor 0.1.6

This project was tested on Ubuntu 18.04 and Windows 10 while the optimization framework requires Linux.

Structure

├── contrllers
│   ├── my_controller_python
│   └── ...
├── advisor_config
│   ├── config.json
│   ├── min_function.py
│   └── ...
├── worlds
│   ├── world_jump.wbt
│   ├── world_pure.wbt
│   └── ...
├── utils
│   ├── plot_result.py
│   └── ...
├── README.md
├── requirements.txt
├── robot_specifications.txt
└── ...

Qucik Start

Simulation Setup

1. Open <worlds/world_pure.wbt>

2. Control the robot using keyboard:

   Motion	Command
   Jump	Space
   Forward	W
   Backward	S
   Left Turn	A
   Right Turn	D
   Stand Up	↑
   Squat Down	↓

Optimization Setup

1. Open <worlds/world_jump.wbt>
2. Check <advisor_config/quick_start.md>, and results can be plotted using <utils/plot_results.py>.