Required enverionments:
- Python 3.6 +
- CoppeliaSim 4.0 +
This project is about the models and algorithms for bipedal walking control, including the Linear Inverted Pendunum Model (LIPM) and humandoid robots. The LIPM is shown as follows:
The calculation for LIPM is implemented with Python 3.6, given different target orbital energy, the LIPM can be controlled to different walking states.
1.1 Giving a initial state for a positve orbital energy.
1.2 Giving a initial state for a negetive orbital energy.
1.2.1 Giving a zero target orbital energy.
1.2.2 Giving a positive target orbital energy.
1.2.3 Giving a negetive target orbital energy.
Run the script in 'LIPM/LIPM_ik_test.py', you can get the following result:
Open 'scenes/demo_1_LIPM_kinematics.ttt' with CoppliaSim 4.0+ and run the project with the begin button of CoppeliaSim , then run with the script 'demo_1_LIPM_kinematics.py', you can get the following result:
Run the script in 'LIPM/demo_LIPM_3D.py', and specify different parameters, you can get different simulation results.
Walk forward:
Turn left:
Turn right:
The code is opensource at: https://github.com/chauby/ZMP_preview_control
Tutorial: https://zhuanlan.zhihu.com/p/452704228?
The ZMP and COM trajectory generation with the given ZMP stepping positions:
代码对应的文章教程有两种访问方式,详见:
-
微信公众号:沙漏AI机器人。
获取《仿人机器人》中文版书籍,可关注公众号后,在后台回复【HR】获取下载链接。