HumanoidStandup
State and action spaces
The default 348-value observation has the same groups and switches as Humanoid: qpos[2..], qvel, non-world cinert, non-world cvel, actuator generalized forces after the root DOFs, and non-world external forces. The default action is 17 controls in \([-0.4,0.4]\).
Dynamics, reward, and info
Let \(z\) be root height, \(h\) the MuJoCo model timestep, \(u\) the simulator controls, and \(f_c\) external contact forces:
\[ r=w_h\frac{z}{h}-w_u\lVert u\rVert_2^2- \operatorname{clip}(w_i\lVert f_c\rVert_2^2,i_{min},i_{max})+1. \]
Info includes x/y position, distance from origin, summed tendon length/velocity, and upright/control/impact reward components.
Episode end
There is no state termination. The horizon truncates.
Configuration
| Field | Default | Meaning |
|---|---|---|
xml_file, frame_skip | embedded, 5 | Model and action duration |
uph_cost_weight | 1 | Upright reward weight |
ctrl_cost_weight | 0.1 | Control penalty |
impact_cost_weight/range | 5e-7, (-∞,10] | Impact penalty and clamp |
reset_noise_scale | 0.01 | Uniform qpos/qvel noise |
| exclude x/y | true | Observation root-position switch |
include cinert, cvel, qfrc_actuator, cfrc_ext | all true | Auxiliary observation groups |
max_episode_steps | 1000 | Horizon |
Reference: Gymnasium HumanoidStandup.