专栏首页CreateAMindSim-to-Real: 仿真训练直接迁移到真实机器人

Sim-to-Real: 仿真训练直接迁移到真实机器人

Sim-to-Real: Learning Agile Locomotion For Quadruped Robots

Jie Tan, Tingnan Zhang, Erwin Coumans, Atil Iscen, Yunfei Bai, Danijar Hafner, Steven Bohez, Vincent Vanhoucke

(Submitted on 27 Apr 2018)

Designing agile locomotion for quadruped robots often requires extensive expertise and tedious manual tuning. In this paper, we present a system to automate this process by leveraging deep reinforcement learning techniques. Our system can learn quadruped locomotion from scratch using simple reward signals. In addition, users can provide an open loop reference to guide the learning process when more control over the learned gait is needed. The control policies are learned in a physics simulator and then deployed on real robots. In robotics, policies trained in simulation often do not transfer to the real world. We narrow this reality gap by improving the physics simulator and learning robust policies. We improve the simulation using system identification, developing an accurate actuator model and simulating latency. We learn robust controllers by randomizing the physical environments, adding perturbations and designing a compact observation space. We evaluate our system on two agile locomotion gaits: trotting and galloping. After learning in simulation, a quadruped robot can successfully perform both gaits in the real world.

https://zhuanlan.zhihu.com/p/36322095

本文分享自微信公众号 - CreateAMind(createamind)

原文出处及转载信息见文内详细说明,如有侵权,请联系 yunjia_community@tencent.com 删除。

原始发表时间:2018-05-07

本文参与腾讯云自媒体分享计划,欢迎正在阅读的你也加入,一起分享。

我来说两句

0 条评论
登录 后参与评论

相关文章

  • 金句频频:用信息瓶颈的迁移学习和探索;关键状态

    We present a hierarchical reinforcement learning (HRL) or options framework for ...

    用户1908973
  • 用信息瓶颈的迁移学习和探索

    Transfer and Exploration via the Information Bottleneck

    用户1908973
  • stackGAN通过文字描述生成图片的V2项目

    https://github.com/hanzhanggit/StackGAN-v2

    用户1908973
  • PHOTON——用于快速机器学习模型开发的Python API(CS-LG)

    本文介绍PHOTON的实现和使用,PHOTON是一个高级的Python API,旨在简化和加速机器学习模型的开发过程。它可以设计基本的和高级的机器学习流水线结构...

    Elva
  • 机器人对话和导航任务的学习和推理(cs.AI)

    强化学习和概率推理算法旨在分别从互动体验和概率语境知识中学习推理。在本研究中,我们开发了机器人任务完成算法,同时研究了强化学习和概率推理技术的辅助优势。机器人从...

    Donuts_choco
  • JavaScript ES6 — 少即是多,以少胜多,四两拨千斤

    JavaScript ES6 brings new syntax and new awesome features to make your code more...

    一个会写诗的程序员
  • 金句频频:用信息瓶颈的迁移学习和探索;关键状态

    We present a hierarchical reinforcement learning (HRL) or options framework for ...

    用户1908973
  • 【最新解读】Ray Dalio——中美之间的误解、争议和战争

    This evolutionary cycle is not just for people but for countries, companies, eco...

    量化投资与机器学习微信公众号
  • 【论文推荐】最新5篇聊天机器人(Chatbot)相关论文—深度强化学习、社交聊天机器人小冰、对话聊天助手、序列-序列、动态词汇

    【导读】专知内容组整理了最近五篇聊天机器人(Chatbot)相关文章,为大家进行介绍,欢迎查看! 1. A Deep Reinforcement Learnin...

    WZEARW
  • ROS机器人项目开发11例-ROS Robotics Projects(6)Matlab和Android

    书中,第8章主要介绍了ROS与Matlab和Android的接口,以及集成使用的方法。

    zhangrelay

扫码关注云+社区

领取腾讯云代金券