Getting Started with Quadruped Robots
Getting Started with Quadruped Robots
Huayang Yin四足入门该学什么
这个页面是给准备学习四足知识的学弟学妹们准备的
有些地方只是我自己的理解，可能会有错误，希望大家及时指正，谢谢！
常见平台
常见的四足机器人
我们实验室使用的四足
理论基础
控制方法
- Model Predictive Control(MPC)
- Trajectory Optimization(TO)
- Reinforcement Learning(RL)
一些网课
- Hands on DL（B站可以搜动手学深度学习）：从MLP到Transformer，从理论到pytorch实现，比较全面的教程
- Optimal Control and RL（CMU）：CMU的课程录像，可以了解一些进阶算法比如MPC等等,youtube网址需要科学上网
- Underactuated Robotics(欠驱动机器人MIT)：MIT的课程录像，可以了解一下
- ROS入门：做机器人必学，实验室用的机器人大部分是ROS控制，不会ROS可以重开了
paper
- Learning to Walk in Minutes Using Massively Parallel Deep Reinforcement Learning（CoRL 2021)
- RMA: Rapid Motor Adaptation for Legged Robots（RSS 2021)
- Safe Reinforcement Learning for Legged Locomotion(IROS 2022)
- Walk These Ways: Tuning Robot Control for Generalization with Multiplicity of Behavior(CoRL 2022)
- Extreme Parkour with Legged Robots(ICRA 2024)
- Sim-to-Real: Learning Agile Locomotion For Quadruped Robots
- Learning quadrupedal locomotion over challenging terrain(Science Robotics 2020)
- A Unified MPC Framework for Whole-Body Dynamic Locomotion and Manipulation(RAL 2021)
- Deep Whole-Body Control: Learning a Unified Policy for Manipulation and Locomotion(CoRL 2022)
- Agile But Safe: Learning Collision-Free High-Speed Legged Locomotion(RSS 2024)
- DTC: Deep Tracking Control(Science Robotics 2024)
- UMI on Legs: Making Manipulation Policies Mobile with Manipulation-Centric Whole-body Controllers(CoRL 2024)
- ViPlanner: Visual Semantic Imperative Learning for Local Navigation
- VLFM: Vision-Language Frontier Maps for Zero-Shot Semantic Navigation(ICRA 2024)
- DreamWaQ: Learning Robust Quadrupedal Locomotion With Implicit Terrain Imagination via Deep Reinforcement Learning(ICRA 2023)
- Hybrid Internal Model: Learning Agile Legged Locomotion with Simulated Robot Response(ICLR 2024)
- Learning Visual Parkour from Generated Images(CoRL 2024)
- Full-Order Sampling-Based MPC for Torque-Level Locomotion Control via Diffusion-Style Annealing
- RoboDuet: A Framework Affording Mobile-Manipulation and Cross-Embodiment
- Playful DoggyBot: Learning Agile and Precise Quadrupedal Locomotion
一些比较好用的仓库
仿真平台
- IsaacGym:这个是四足RL最常用的平台，如果后续要开发的话，建议详细地看一下Nvidia官方的文档
- IsaacSim:对于配置要求较高，没系统地用过，后面打算把工作迁移到这里
- mujoco:没系统地使用过
- Genesis:真实性有待考察
算法
- Curobo:前两天用这个算机械臂的逆运动学来着，并行，又快又准
