V
主页
Consensus Complementarity Control for Multi-Contact MPC
发布人
https://www.youtube.com/watch?v=L57Jz3dPwO8 Consensus Complementarity Control for Multi-Contact MPC Alp Aydinoglu, Adam Wei, Wei-Cheng Huang, Michael Posa We propose a hybrid model predictive control algorithm, consensus complementarity control (C3), for systems that make and break contact with their environment. Many state-of-the-art controllers for tasks which require initiating contact with the environment, such as locomotion and manipulation, require a priori mode schedules or are too computationally complex to run at real-time rates. We present a method based on the alternating direction method of multipliers (ADMM) that is capable of high-speed reasoning over potential contact events. Via a consensus formulation, our approach enables parallelization of the contact scheduling problem. We validate our results on five numerical examples, including four high-dimensional frictional contact problems, and a physical experimentation on an underactuated multi-contact system. We further demonstrate the effectiveness of our method on a physical experiment accomplishing a high-dimensional, multi-contact manipulation task with a robot arm. https://arxiv.org/abs/2304.11259
打开封面
下载高清视频
观看高清视频
视频下载器
AI 编程工具 Cursor 带来哪些机遇与挑战(20240908直播回放)
基于仿真系统的强化学习系列,pybullet仿真系统篇(1)
模型预测控制 (MPC) 从推导到C++ 实现,详细教程 仅使用 Eigen 库 代码在简介
MPC and RL, two different roads to legged locomotion, and that's OK
Tailoring model complexity in MPC of legged locomotion
波士顿动力 MPC 控制策略的演进与思考
MuJoCo 仿真 案例欣赏
北京机器人展迷惑事件
令人惊叹的步态和控制 无比丝滑 Boston Dynamics Spot
通过优化理论指导设计双足以达到控制和任务的最优
Part 1 - How to Solve Inverse Kinematics of a 4 Leg Robot
Inverse-Dynamics MPC via Nullspace Resolution (IROS'23 Presentation)
人形机器人远程视觉操纵 论文和代码在简介
Contact-Implicit MPC: Controlling Diverse Quadruped Motions Without Pre-Planned
[23 RAL]利用实时非线性 MPC 实现双足运动的无缝反应策略
Tutorial - Crocoddyl - Nicolas Mansard
有开源github代码! RA-L论文 人形机器人的多接触全身控制
人形机器人足底机械结构设计
【强化学习训练走路的机器人】油管最强动画强化学习入门!不愧是MATLAB官方教程,行走机器人|深度学习|AI 人工智能
PACC:使用 MPC 的四足机器人高负载协作搬运被动臂方法
人形机器人远程视觉操纵
[下]6个人形双足强化学习开源项目,论文讲解,代码速读,FLD,PBRS,footstep,ExBody,humanplus,humanoid-gym
Vision-Based Terrain-Aware Locomotion for Legged Robots ---T-RO
Curiosity-Driven Learning of Joint Locomotion and Manipulation Tasks
Real-time Model Predictive Control with Zonotope-Based Neural Networks for Biped
华北理工大学ROS2智能机器人算法 实验一
【深度强化学习】华盛顿大学Steve Brunton教授精讲《数据驱动的科学与工程:机器学习、动态系统与控制》
[IROS 2024] 通过分布式优化加速腿式机器人的模型预测控制
动态变化环境中单足跳跃者的模型预测跑酷控制 代码地址在简介
Learning Morphologically Adaptive Locomotion Controllerfor Quadruped
人形机器人腿部机械结构设计 !很简单,有手就行!论文在简介
Topology-Based MPC for Automatic Footstep Placement and Contact Surface Selectio
人形双足模仿学习强化学习AMP ASE Exbody求助
(Humanoid)An IntegratedForceSensitive, Impedance Controlled, Tendon-Driven Wrist
Foothold Evaluation Criterion for Dynamic Transition Feasibility for Quadruped
PID超大狗制作教程 [更到5集]
OpenRLHF:大规模分布式RLHF训练系统介绍
Sim-to-Real Learning for Bipedal Locomotion Under Unsensed Dynamic Loads
Combining Learning-Based Locomotion with Model-Based Manipulation for Legged
(ICRA 2024)Contrastive Initial State Buffer for Reinforcement Learning