Page 101 - Read Online

Zhang et al. Intell Robot 2022;2(3):275­97 I http://dx.doi.org/10.20517/ir.2022.20 Page 291

ANYmal-B, ANYmal-C Jueying 3 Jueying-Mini robot ANYmal Panther Black robot A1 Unitree Unitree Laikago A1 Unitree

RaiSim PyBullet RaiSim Mujoco RaiSim Pybullet Pybullet Pybullet

Velocity, and Target Velocity, Veloc- Ref- State, Contacts, Torque Position unifor- Pitch and Position, Base Lin- and Velocity, Contact. and Joint Tar- to

Angular Reward Foot Clearance, Torque. and and Height (Torque, Body and Joint Gait, Torque Base Velocities, Desired Foot and Velocities, Velocity, Torque. and Velocity

and Motion Pose, Regularisation State, Foot Positions Position. Velocity, Velocity, Smoothness, Positions, End-Effector Positions, Angular Torque, Balance, Base Forward Count, Forward Orientation.
Linear Base Collision, Smoothness, Base ity), erence Goal Base and uniformity, mity, Limitations. Joint and ear Quaternion. Joint Base Torques, Survival. Base Limit Desired and get

and Residu- Joint Gating: Joint Joint Joint Fre- Cutoff Stance Offset. Joint Base Velocity,
Frequencies Position Desired Weights. Leg and Phase Swing, and

Leg Foot als. Expert: Positions. Variable Desired Positions. Desired Positions. Desired Positions. Desired quency, (Swing Phase), Desired Torques. Goal Height Orientation.

Base Joint FTG Joint Vector, and States, Ve- Base Goal Direction, Lin- Posi- (Base, Ac- Linear Base An- Po- Gait

Gravity, Frequency, Velocities, Frequencies, Normal Forces Contact Contact Force. Gravity, and Vector, Gravity and Velocity Joint and Velocity. End-Effector, Velocities Previous Base Height, Joint and Motors Map, and

Direction, and and and Terrain Height, History, External Position, Phase Height, Angular Acceleration, Angular (Base, State CoM), Orientation, Command. VelocityCommand,SineandCo- sine Values (4 phases), Joint Posi- tion and Velocity, Angular Veloc- Gravity. Actual and Base Velocities (12-dim). Height Orientation,

Goal Velocities States Phases History, Foot Target Friction, Joint locities, Position. Base Base ear and tion Image, Joint, Joint), tion, ities, Desired Velocity Orientation, Linear gles Global sitions, Phase.

TRPO SAC PPO2 V-MPO [25] , MO- VMPO [26] , PPO ARS [23] PPO SAC

incor- show- gener- generate learning combines ob- re- to framework frame- transitions re- a adapta- trajectories and

solution proprioception zero-shot to for which DRL. controller policies gaits. learning gait with domain high-level

Sim2Real remarkable architecture adaptiveskillsfromagroupofex- method gaits, and adaptive training producetrajectoriesplannedbya solver. quadrupedal for training a control policy to lo- various in hierarchical which automatically energy. min. for tion by identifying a simulator to simulated ones. target AnovelHTCframeworkleverag- the for optimal control for the low-level.

novel porating alization. MEL skills. training bounding pre-training terrain by tained non-linear novel comote in emerge of framework the match the DRL

A ing A pert A A A A work ward A to ing

2020 2020 2021 2021 and 2021 2021 2021 2021
Science Robotics Science Robotics ArXiv ArXiv IEEE Robotics Automation Letters CoRL ICRA IROS

Locomotion adap- of (MEL) Policies Control us- Bounding Networks [57] Terrain- a Imitating Planner [27] for Transition Phase-Guided via Locomotion Control Locomotion

Quadrupedal Terrain [7] Challenging learning locomotion [9] of Learning Quadruped Neural Coordinated by Locomotion Dynamics Gait Free via Robots Efficient Transitions [52] SimGAN: Hybrid Simulator Identifica- tionforDomainAdaptationviaAdver- Terrain-Aware Quadrupedal by Combining DRL and Optimal Con- 3 https://www.deeprobotics.cn/

Learning over Multi-expert legged tive Efficient Robust for Pretrained ing Learning Adaptive Centroidal Learning Quadruped Controller [58] and Fast Gait Learned RL [104] sarial Hierarchical for (HTC) trol [54]