Skip to content

Latest commit

 

History

History
95 lines (64 loc) · 3.35 KB

README.md

File metadata and controls

95 lines (64 loc) · 3.35 KB

nabo_pointfoot

"Naughty boy Pointfoot" 🤖

GitHub stars

Bilibili | Nabo | 简体中文


demo.mp4

Introduction

  • Nabo Pointfoot is a pointfoot bipedal robot simulation framework in MuJoCo, constructed fully with C++.

  • Nabo Pointfoot is built on top of nabo, which is a bipedal robot framework for a 12DOF robot model.

  • MPC+WBIC balence controller. The controller code is isolated from simulation, which is elegant.

  • Nabo Pointfoot walks with the height of 0.45m, maximux walking speed 0.95m/s.


Features

  • 💻Pure C++
  • ⚡️Computing Efficient, formulating QP with eiquadprog
  • 🤹Interactive MuJoCo GUI
  • 📝Rich code comments but most in Chinese, I apologize for those who don't understand Chinese

nabo_pointfoot

Size Comparison with LIMX P1 & Unitree H1 🤖


Usage

  1. This project was developed on Ubuntu20.04, g++/gcc version 9.4.0.
  2. $ git clone https://github.com/SeaHI-Robot/Nabo_Pointfoot_Bipedal_Robot.git, then cd to this repo's directory.
  3. $ ./make_and_run to compile both the controller and the simulation, then start simulation with the compiled controller loaded.
  4. When both codes in nabo_core/ and nabo_mujoco/ are compiled,simply do $ ./run_sim.sh to only start the simulation without any compilation.
  5. Edit nabo_mujoco/000.ini to adjust the parameters for tuning. nabo_mujoco/zzz.txt stores the log file during simulation when the log flag is turned on.
  6. Pitch is currently poorly controlled at close to zero speed and will stabilize with a bias. Stable yaw angle control cannot be achieved for the time being.

In the MuJoCo GUI:

Press:

  • w/s : change command speed in x
  • a/d : change command speed in y
  • j/l : change command angular speed in z
  • x: clear all speed command
  • backspace: reset
  • space: pause
  • f: toggle visualization of contact force
  • z: toggle log flag

Discussion

If U have any problem, please start an Issue, or join QQ group 609601974.


Acknowledgement

  1. nabo,

  2. MIT mini-cheetah,

  3. MuJoCo Simulator,

  4. and SUSTech🌈.

SUSTech