-
Notifications
You must be signed in to change notification settings - Fork 0
/
single_actor.py
150 lines (115 loc) · 4.45 KB
/
single_actor.py
1
2
3
4
5
6
7
8
9
10
11
12
13
14
15
16
17
18
19
20
21
22
23
24
25
26
27
28
29
30
31
32
33
34
35
36
37
38
39
40
41
42
43
44
45
46
47
48
49
50
51
52
53
54
55
56
57
58
59
60
61
62
63
64
65
66
67
68
69
70
71
72
73
74
75
76
77
78
79
80
81
82
83
84
85
86
87
88
89
90
91
92
93
94
95
96
97
98
99
100
101
102
103
104
105
106
107
108
109
110
111
112
113
114
115
116
117
118
119
120
121
122
123
124
125
126
127
128
129
130
131
132
133
134
135
136
137
138
139
140
141
142
143
144
145
146
147
148
149
150
from logging import root
import math
import random
from turtle import up
import numpy as np
from isaacgym import gymapi, gymutil, gymtorch
import torch
# simulation setup
gym = gymapi.acquire_gym()
# Creating a Simulation
sim_params = gymapi.SimParams()
sim_params.dt = dt = 1.0 / 60.0
sim_params.physx.solver_type = 1
sim_params.physx.num_position_iterations = 6
sim_params.physx.num_velocity_iterations = 0
sim_params.physx.num_threads = 0
sim_params.physx.use_gpu = True
# activate use_gpu_pipeline to enable Tensor API
sim_params.use_gpu_pipeline = True
# uses PhysX backend
sim = gym.create_sim(0, 0, gymapi.SIM_PHYSX, sim_params)
# configure the ground plane
plane_params = gymapi.PlaneParams()
gym.add_ground(sim, plane_params)
# creating a viewer
viewer = gym.create_viewer(sim, gymapi.CameraProperties())
# load the humanoid with tennis racket mjcf file
asset_root = "/home/brandon/isaacgym/assets"
asset_file = "mjcf/nv_humanoid.xml"
# asset options
asset_options = gymapi.AssetOptions()
asset_options.fix_base_link = True
asset_options.flip_visual_attachments = False
asset_options.use_mesh_materials = True
# load asset function
asset = gym.load_asset(sim, asset_root, asset_file, asset_options)
# set up the env grid
num_envs = 1
num_per_row = 1
spacing = 2.5
env_lower = gymapi.Vec3(-spacing, 0.0, -spacing)
env_upper = gymapi.Vec3(spacing, spacing, spacing)
# position the camera
# cam_pos = gymapi.Vec3(17.2, 2.0, 16)
# cam_target = gymapi.Vec3(5, -2.5, 13)
# gym.viewer_camera_look_at(viewer, None, cam_pos, cam_target)
# cashe some common handles for later use
envs = []
actor_handles = []
print("Creating %d environment" % num_envs)
for i in range(num_envs):
env = gym.create_env(sim, env_lower, env_upper, num_per_row)
envs.append(env)
pose = gymapi.Transform()
pose.p = gymapi.Vec3(0.0, 1.32, 0.0)
pose.r = gymapi.Quat(-0.707107, 0.0, 0.0, 0.707107)
actor_handle = gym.create_actor(env, asset, pose, "actor", i, 1)
actor_handles.append(actor_handle)
num_bodies = gym.get_actor_rigid_body_count(env, actor_handle)
num_joints = gym.get_actor_joint_count(env, actor_handle)
num_dofs = gym.get_actor_dof_count(env, actor_handle)
print("number of bodies: ", num_bodies)
print("number of joints: ", num_joints)
print("number of dofs: ", num_dofs)
gym.prepare_sim(sim)
# Actor Root State Tensor
_root_tensor = gym.acquire_actor_root_state_tensor(sim)
# wrap in PyTorch Tensor
root_tensor = gymtorch.wrap_tensor(_root_tensor)
# size = (num_actors, the number of elements)
root_positions = root_tensor[:, 0:3]
root_orientations = root_tensor[:, 3:7]
root_linvels = root_tensor[:, 7:10]
root_angvels = root_tensor[:, 10:13]
torch.set_printoptions(profile="full")
# size of rb_states: (num_rigid_bodies, 13), (17, 13) in this case
# num_rigid_bodies = num_bodies * num_actors
# 13 is: 3 position + 4 orientation + 3 linear velociy + 3 angular velocity
_rb_states = gym.acquire_rigid_body_state_tensor(sim)
rb_states = gymtorch.wrap_tensor(_rb_states)
# dof_states = position, velocity
_dof_states = gym.acquire_dof_state_tensor(sim)
dof_states = gymtorch.wrap_tensor(_dof_states)
# get lower and upper bounds of dofs.
props = gym.get_actor_dof_properties(env, actor_handle)
lower_bound_list = props["lower"]
upper_bound_list = props["upper"]
while not gym.query_viewer_has_closed(viewer):
# step the physics
gym.simulate(sim)
gym.fetch_results(sim, True)
# refresh
gym.refresh_actor_root_state_tensor(sim)
gym.refresh_rigid_body_state_tensor(sim)
gym.refresh_dof_state_tensor(sim)
# compute velocity of the "racket head"
# 13 is the index of the racket; rigid body 7:10 are the index for xyz velocities respecitvely
xyz_velocity = rb_states[13][7:10]
speed = math.sqrt(xyz_velocity[0].item()**2 + xyz_velocity[1].item()**2 + xyz_velocity[2].item()**2)
# position = (3 position, 4 orientation)
position = rb_states[:, 0:7]
# perform random action
for i in range(len(dof_states)):
random_action = random.uniform(lower_bound_list[i], upper_bound_list[i])
dof_states[i] = torch.tensor([random_action, 0], device="cuda:0")
# applies all the values in the tensor (reverse kinematics)
gym.set_dof_state_tensor(sim, _dof_states)
# update the viewer
gym.step_graphics(sim)
gym.draw_viewer(viewer, sim, True)
# Wait for dt to elapse in real time.
# This synchronizes the physics simulation with the rendering rate.
gym.sync_frame_time(sim)
gym.destroy_viewer(viewer)
gym.destory_sim(sim)