jaykorea/Isaac-RL-Two-wheel-Legged-Bot：IsaacLab 的两轮机器人——强化学习

352

MIT License

项目介绍

jaykorea/Isaac-RL-Two-wheel-Legged-Bot 是一个托管在 GitHub 上的开源项目，主要使用 Python 编写，采用 MIT License 许可证，目前获得 352 个 Star、26 次 Fork。

仓库全名: jaykorea/Isaac-RL-Two-wheel-Legged-Bot
开源许可证: MIT License
Star 数: 352
Fork 数: 26
未关闭 Issue: 6

README 摘要

Isaac LAB for Flamingo ✨ New Features - Updated 🚀 ✔️ Flamingo rev.0.1.4: Latest version of Flamingo added. ✔️ Flamingo Edu v1: Flamingo Edu version added. ✔️ Stack Environment: Observations can be stacked with arguments. ✔️ Constraint Manager: Constraints as Termination (CaT) method implementation added. ✔️ CoRL: Based on rslrl library, off-policy algorithms are implemented on offpolicyrunner. Sim2Real - ZeroShot…

完整内容见 GitHub 原始 README 或本页。

常见问题

Isaac-RL-Two-wheel-Legged-Bot 用什么语言开发？

主要使用 Python 编写，采用 MIT License 许可证，目前有 352 个 Star。

Isaac-RL-Two-wheel-Legged-Bot 可以免费商用吗？

如何获取 Isaac-RL-Two-wheel-Legged-Bot 的源码？

可以直接克隆仓库：git clone https://github.com/jaykorea/Isaac-RL-Two-wheel-Legged-Bot.git，默认分支为 main；安装与运行步骤见仓库 README。

Isaac-RL-Two-wheel-Legged-Bot 最近还在维护吗？

仓库最近一次更新为 3月26日，当前有 6 个未关闭 Issue；Star 数为 352，可作为社区活跃度的参考。

关于

GitHub 原始描述（英文）

two wheel legged robot for IsaacLab - reinforcement learning

描述取自 GitHub 官方 API，保持仓库作者书写的原文。

简要分析

热度分

414

价值分

219

活跃状态

活跃

bipedal-robots isaac-gym isaac-sim isaaclab reinforcement-learning two-wheel-robot

GitHub

README

Isaac LAB for Flamingo

✨ New Features - Updated 🚀

✔️ Flamingo rev.0.1.4: Latest version of Flamingo added.
✔️ Flamingo Edu v1: Flamingo Edu version added.
✔️ Stack Environment: Observations can be stacked with arguments.
✔️ Constraint Manager: Constraints as Termination (CaT) method implementation added.
✔️ CoRL: Based on rsl_rl library, off-policy algorithms are implemented on off_policy_runner.

Sim2Real - ZeroShot Transfer

Isaac Lab Flamingo

Sim 2 Sim framework - Lab to MuJoCo

Simulation to Simulation framework is available on sim2sim_onnx branch (Currently on migration update)
You can simply inference trained policy (basically export as .onnx from isaac lab)

Setup

This repo is tested on Ubuntu 20.04, and I recommend you to install ‘local install’

1. Install Isaac Sim

https://isaac-sim.github.io/IsaacLab/main/source/setup/installation/binaries_installation.html

2. Install Isaac Lab

https://github.com/isaac-sim/IsaacLab

3. Install lab.flamingo package

i. clone repository

git clone https://github.com/jaykorea/Isaac-RL-Two-wheel-Legged-Bot

ii. install lab.flamingo pip package by running below command

run it on ‘lab.flamingo’ root path

conda activate env_isaaclab # change to you conda env
pip install -e .

iii. Unzip assets(usd asset) on folder

Since git does not correctly upload ‘.usd’ file, you should manually unzip the usd files on assests folder

 path example: lab/flamingo/assets/data/Robots/Flamingo/flamingo_rev01_4_1/

Launch script

Train flamingo

run it on ‘lab.flamingo’ root path

  python scripts/co_rl/train.py --task {task name} --algo ppo --num_envs 4096 --headless --num_policy_stacks {stack number on policy obs} --num_critic_stacks {stack number on critic obs}

Train example - track velocity

  python scripts/co_rl/train.py --task Isaac-Velocity-Flat-Flamingo-v1-ppo --algo ppo --num_envs 4096 --headless --num_policy_stacks 2 --num_critic_stacks 2

play flamingo

run it on ‘lab.flamingo’ root path

  python scripts/co_rl/play.py --task {task name} --algo ppo --num_envs 64 --num_policy_stacks {stack number on policy obs} --num_critic_stacks {stack number on critic obs} --load_run {folder name} --plot False

play example - track velocity

  python scripts/co_rl/play.py --task Isaac-Velocity-Flat-Flamingo-Play-v1-ppo --algo ppo --num_envs 64 --num_policy_stacks {stack number on policy obs} --num_critic_stacks {stack number on critic obs} --load_run 2025-03-16_17-09-35 --plot False

Isaac-RL-Two-wheel-Legged-Bot