Page 98 - Read Online
P. 98
Page 290
Spot
Hexapod
ANYmal
Unitree-
Unitree
Laikago
Unitree
Laikago
Daisy
Mini
IsaacGym [102]
and
MuJoCo.
Pybullet
PyBullet
PyBullet
Pybullet
RaiSim,
PyBullet
Tar-
Velocity,
Orienta-
Velocities,
Velocity,
State,
Refer-
(2-dim),
and
(3-dim).
Acceleration
Base
Foot
Velocity.
to
Speed,
and
Smoothness,
vs.
(Base
Positions,
Velocities
and
Linear
Orientation
Velocities
Angular
(Current
Base
Linear
Joint
Distance
Velocity).
Distance
Foot
Deviation.
Base
Torque,
Angular
End-Effector
Locomotion:
Base
Joint
Gap
Orientation.
and
Slip,
Euclidean
Distance,
and
Forward
Torques,
Desired
Motion
Torque,
in
Linear
Cost,
Pose
Base
ence
tion
and
De-
Residu-
Frequency,
Primitive
Joint
Joint
Vector
Stride
for
Positions.
Height,
Torques
Position
Positions.
Positions.
Selection
One-Hot
Desired
Desired
(9-dim)
Action.
Swing
sired
Foot
Joint
Gait
als.
Pre-
Tar-
Se-
Out-
Ori-
Orientation,
Linear
Foot
and
Primitive
positions
Action
Direction,
Angles,
Velocities,
(3-dim)
Policy
and
Velocity.
and
(3-dim),
linear
positions),
(3-dim)
and
State,
Joint
Env.,
(Orientation
(previously-used).
Base
and
and
Rotations)
(without
Joint
(1-dim)
Velocity,
Accelerations
Velocities
Distance
to
Desired
Height
Param.
foot
entation,
Distance
quence.
Height, Recov- Fluency. Direction, State, Joint Velocities, Base Joint Desired ANYmal RaiSim HAA, Motion, Joint Torque, ery: State(withHistory),PreviousAc- Positions. In- Impulse, and Slip KFE, HFE, Gravity. Command, tion, Flu- Direction, Contact, ternal ency. IMU (6-step), Angles Motor and Yaw, and Position Base Joint Desired Minitaur Pybullet Previous and (6-step), Readings Smoothness. Positions. (6-step). Action Angu- Base (2-dim), Orientation Zhang et al. Intell Robot 2022;2(3):27597 Laikago Duration. Timestep and get) Residual Length, Genera- Trajectory Action,
Phase.
Phase
Joint
Base
Base
Pose
Pose
and
put,
tor
get
lar
TRPO SAC ARS [23] GCPO DQN PPO SAC PPO
lever- and generation learning locomotion the in human frame- dy- with randomized ter- framework commands com- Control robust ter- for ground-truth generaliz- for on skills robot skills animals.
method automated data for RL Deep minimal Sim2Real domain uneven traversing RL velocity framework Model-Based synthesize controllers. approach rain locomotion using exterocep- without and framework locomotion legged locomotion real-world
Sim2Real fast, cost-effective system quadrupedal with with world quadrupedal work utilizing offline RL and allow CPPO-based tracking constraints. hierarchical to RL quadrupedal learning inputs maps. sample-efficient hierarchical robots. real-world enabling system agile learn imitating
A aging schemes. A policies real effort. A namics to rain. A for under A bining and A tive height A able learning A to by
2019 2020 2020 2020 2020 2020 2020
Science Robotics CoRL ArXiv IEEE Robotics Autom CoRL ArXiv ICRA 2020 RSS
motor World Random- Bezier Legged Optimiza- Robot Con- Lo- Legged for Locomotion Reinforce- Locomotion
dynamic Real the Effort [41] Domain with Policy Quadrupedal Contact-Adaptive Efficient Generalization Policies [103] Animals [10]
and robots [6] in Human Modulation Sim-to-Real Generalizable Hierarchical Robotic
agile legged Walk to Minimal and for Constrained Dynamic a Robust, Terrain Locomotion with Learning [47] Agile Imitating
Learning for skills Learning with Dynamics Gait ized Curves Locomotion [50] Guided for tion Locomotion [28] Learning for troller comotion [101] Zero-Shot Visual Learning Skills ment Learning by Skills