Working Beerpong and TableTennis with MPWrapper
This commit is contained in:
Marcel
@@ -0,0 +1 @@
|
||||
from .mp_wrapper import MPWrapper
|
||||
@@ -0,0 +1,187 @@
|
||||
<mujoco model="wam(v1.31)">
|
||||
<compiler angle="radian" meshdir="../../meshes/wam/" />
|
||||
<option timestep="0.005" integrator="Euler" />
|
||||
<size njmax="500" nconmax="100" />
|
||||
<default class="main">
|
||||
<joint limited="true" frictionloss="0.001" damping="0.07"/>
|
||||
<default class="viz">
|
||||
<geom type="mesh" contype="0" conaffinity="0" group="1" rgba="0.7 0.7 0.7 1" />
|
||||
</default>
|
||||
<default class="col">
|
||||
<geom type="mesh" contype="0" rgba="0.5 0.6 0.7 1" />
|
||||
</default>
|
||||
</default>
|
||||
<asset>
|
||||
<texture type="2d" name="groundplane" builtin="checker" mark="edge" rgb1="0.25 0.26 0.25" rgb2="0.22 0.22 0.22" markrgb="0.3 0.3 0.3" width="100" height="100" />
|
||||
<material name="MatGnd" texture="groundplane" texrepeat="5 5" specular="1" shininess="0.3" reflectance="1e-05" />
|
||||
<mesh name="base_link_fine" file="base_link_fine.stl" />
|
||||
<mesh name="base_link_convex" file="base_link_convex.stl" />
|
||||
<mesh name="shoulder_link_fine" file="shoulder_link_fine.stl" />
|
||||
<mesh name="shoulder_link_convex_decomposition_p1" file="shoulder_link_convex_decomposition_p1.stl" />
|
||||
<mesh name="shoulder_link_convex_decomposition_p2" file="shoulder_link_convex_decomposition_p2.stl" />
|
||||
<mesh name="shoulder_link_convex_decomposition_p3" file="shoulder_link_convex_decomposition_p3.stl" />
|
||||
<mesh name="shoulder_pitch_link_fine" file="shoulder_pitch_link_fine.stl" />
|
||||
<mesh name="shoulder_pitch_link_convex" file="shoulder_pitch_link_convex.stl" />
|
||||
<mesh name="upper_arm_link_fine" file="upper_arm_link_fine.stl" />
|
||||
<mesh name="upper_arm_link_convex_decomposition_p1" file="upper_arm_link_convex_decomposition_p1.stl" />
|
||||
<mesh name="upper_arm_link_convex_decomposition_p2" file="upper_arm_link_convex_decomposition_p2.stl" />
|
||||
<mesh name="elbow_link_fine" file="elbow_link_fine.stl" />
|
||||
<mesh name="elbow_link_convex" file="elbow_link_convex.stl" />
|
||||
<mesh name="forearm_link_fine" file="forearm_link_fine.stl" />
|
||||
<mesh name="forearm_link_convex_decomposition_p1" file="forearm_link_convex_decomposition_p1.stl" />
|
||||
<mesh name="forearm_link_convex_decomposition_p2" file="forearm_link_convex_decomposition_p2.stl" />
|
||||
<mesh name="wrist_yaw_link_fine" file="wrist_yaw_link_fine.stl" />
|
||||
<mesh name="wrist_yaw_link_convex_decomposition_p1" file="wrist_yaw_link_convex_decomposition_p1.stl" />
|
||||
<mesh name="wrist_yaw_link_convex_decomposition_p2" file="wrist_yaw_link_convex_decomposition_p2.stl" />
|
||||
<mesh name="wrist_pitch_link_fine" file="wrist_pitch_link_fine.stl" />
|
||||
<mesh name="wrist_pitch_link_convex_decomposition_p1" file="wrist_pitch_link_convex_decomposition_p1.stl" />
|
||||
<mesh name="wrist_pitch_link_convex_decomposition_p2" file="wrist_pitch_link_convex_decomposition_p2.stl" />
|
||||
<mesh name="wrist_pitch_link_convex_decomposition_p3" file="wrist_pitch_link_convex_decomposition_p3.stl" />
|
||||
<mesh name="wrist_palm_link_fine" file="wrist_palm_link_fine.stl" />
|
||||
<mesh name="wrist_palm_link_convex" file="wrist_palm_link_convex.stl" />
|
||||
<mesh name="cup1" file="cup_split1.stl" scale="0.001 0.001 0.001" />
|
||||
<mesh name="cup2" file="cup_split2.stl" scale="0.001 0.001 0.001" />
|
||||
<mesh name="cup3" file="cup_split3.stl" scale="0.001 0.001 0.001" />
|
||||
<mesh name="cup4" file="cup_split4.stl" scale="0.001 0.001 0.001" />
|
||||
<mesh name="cup5" file="cup_split5.stl" scale="0.001 0.001 0.001" />
|
||||
<mesh name="cup6" file="cup_split6.stl" scale="0.001 0.001 0.001" />
|
||||
<mesh name="cup7" file="cup_split7.stl" scale="0.001 0.001 0.001" />
|
||||
<mesh name="cup8" file="cup_split8.stl" scale="0.001 0.001 0.001" />
|
||||
<mesh name="cup9" file="cup_split9.stl" scale="0.001 0.001 0.001" />
|
||||
<mesh name="cup10" file="cup_split10.stl" scale="0.001 0.001 0.001" />
|
||||
<mesh name="cup11" file="cup_split11.stl" scale="0.001 0.001 0.001" />
|
||||
<mesh name="cup12" file="cup_split12.stl" scale="0.001 0.001 0.001" />
|
||||
<mesh name="cup13" file="cup_split13.stl" scale="0.001 0.001 0.001" />
|
||||
<mesh name="cup14" file="cup_split14.stl" scale="0.001 0.001 0.001" />
|
||||
<mesh name="cup15" file="cup_split15.stl" scale="0.001 0.001 0.001" />
|
||||
<mesh name="cup16" file="cup_split16.stl" scale="0.001 0.001 0.001" />
|
||||
<mesh name="cup17" file="cup_split17.stl" scale="0.001 0.001 0.001" />
|
||||
<mesh name="cup18" file="cup_split18.stl" scale="0.001 0.001 0.001" />
|
||||
<mesh name="cup3_table" file="cup_split3.stl" scale="0.00211 0.00211 0.01" />
|
||||
<mesh name="cup4_table" file="cup_split4.stl" scale="0.00211 0.00211 0.01" />
|
||||
<mesh name="cup5_table" file="cup_split5.stl" scale="0.00211 0.00211 0.01" />
|
||||
<mesh name="cup6_table" file="cup_split6.stl" scale="0.00211 0.00211 0.01" />
|
||||
<mesh name="cup7_table" file="cup_split7.stl" scale="0.00211 0.00211 0.01" />
|
||||
<mesh name="cup8_table" file="cup_split8.stl" scale="0.00211 0.00211 0.01" />
|
||||
<mesh name="cup9_table" file="cup_split9.stl" scale="0.00211 0.00211 0.01" />
|
||||
<mesh name="cup10_table" file="cup_split10.stl" scale="0.00211 0.00211 0.01" />
|
||||
<mesh name="cup15_table" file="cup_split15.stl" scale="0.00211 0.00211 0.01" />
|
||||
<mesh name="cup16_table" file="cup_split16.stl" scale="0.00211 0.00211 0.01" />
|
||||
<mesh name="cup17_table" file="cup_split17.stl" scale="0.00211 0.00211 0.01" />
|
||||
<mesh name="cup18_table" file="cup_split18.stl" scale="0.00211 0.00211 0.01" />
|
||||
</asset>
|
||||
|
||||
<worldbody>
|
||||
<geom name="ground" size="5 5 1" type="plane" material="MatGnd" />
|
||||
<light pos="0.1 0.2 1.3" dir="-0.0758098 -0.32162 -0.985527" directional="true" cutoff="60" exponent="1" diffuse="1 1 1" specular="0.1 0.1 0.1" />
|
||||
|
||||
<body name="wam/base_link" pos="0 0 0.6">
|
||||
<inertial pos="6.93764e-06 0.0542887 0.076438" quat="0.496481 0.503509 -0.503703 0.496255" mass="27.5544" diaginertia="0.432537 0.318732 0.219528" />
|
||||
<geom class="viz" quat="0.707107 0 0 -0.707107" mesh="base_link_fine" />
|
||||
<geom class="col" quat="0.707107 0 0 -0.707107" mesh="base_link_convex" name="base_link_convex_geom"/>
|
||||
<body name="wam/shoulder_yaw_link" pos="0 0 0.16" quat="0.707107 0 0 -0.707107">
|
||||
<inertial pos="-0.00443422 -0.00066489 -0.12189" quat="0.999995 0.000984795 0.00270132 0.00136071" mass="10.7677" diaginertia="0.507411 0.462983 0.113271" />
|
||||
<joint name="wam/base_yaw_joint" pos="0 0 0" axis="0 0 1" range="-2.6 2.6" />
|
||||
<geom class="viz" pos="0 0 0.186" mesh="shoulder_link_fine" />
|
||||
<geom class="col" pos="0 0 0.186" mesh="shoulder_link_convex_decomposition_p1" name="shoulder_link_convex_decomposition_p1_geom"/>
|
||||
<geom class="col" pos="0 0 0.186" mesh="shoulder_link_convex_decomposition_p2" name="shoulder_link_convex_decomposition_p2_geom"/>
|
||||
<geom class="col" pos="0 0 0.186" mesh="shoulder_link_convex_decomposition_p3" name="shoulder_link_convex_decomposition_p3_geom"/>
|
||||
<body name="wam/shoulder_pitch_link" pos="0 0 0.184" quat="0.707107 -0.707107 0 0">
|
||||
<inertial pos="-0.00236983 -0.0154211 0.0310561" quat="0.961781 -0.272983 0.0167269 0.0133385" mass="3.87494" diaginertia="0.0214207 0.0167101 0.0126465" />
|
||||
<joint name="wam/shoulder_pitch_joint" pos="0 0 0" axis="0 0 1" range="-1.985 1.985" />
|
||||
<geom class="viz" mesh="shoulder_pitch_link_fine" />
|
||||
<geom class="col" mesh="shoulder_pitch_link_convex" />
|
||||
<body name="wam/upper_arm_link" pos="0 -0.505 0" quat="0.707107 0.707107 0 0">
|
||||
<inertial pos="-0.0382586 3.309e-05 -0.207508" quat="0.705455 0.0381914 0.0383402 0.706686" mass="1.80228" diaginertia="0.0665697 0.0634285 0.00622701" />
|
||||
<joint name="wam/shoulder_yaw_joint" pos="0 0 0" axis="0 0 1" range="-2.8 2.8" />
|
||||
<geom class="viz" pos="0 0 -0.505" mesh="upper_arm_link_fine" />
|
||||
<geom class="col" pos="0 0 -0.505" mesh="upper_arm_link_convex_decomposition_p1" name="upper_arm_link_convex_decomposition_p1_geom"/>
|
||||
<geom class="col" pos="0 0 -0.505" mesh="upper_arm_link_convex_decomposition_p2" name="upper_arm_link_convex_decomposition_p2_geom"/>
|
||||
<body name="wam/forearm_link" pos="0.045 0 0.045" quat="0.707107 -0.707107 0 0">
|
||||
<inertial pos="0.00498512 -0.132717 -0.00022942" quat="0.546303 0.447151 -0.548676 0.447842" mass="2.40017" diaginertia="0.0196896 0.0152225 0.00749914" />
|
||||
<joint name="wam/elbow_pitch_joint" pos="0 0 0" axis="0 0 1" range="-0.9 3.14159" />
|
||||
<geom class="viz" mesh="elbow_link_fine" />
|
||||
<geom class="col" mesh="elbow_link_convex" />
|
||||
<geom class="viz" pos="-0.045 -0.073 0" quat="0.707388 0.706825 0 0" mesh="forearm_link_fine" />
|
||||
<geom class="col" pos="-0.045 -0.073 0" quat="0.707388 0.706825 0 0" mesh="forearm_link_convex_decomposition_p1" name="forearm_link_convex_decomposition_p1_geom" />
|
||||
<geom class="col" pos="-0.045 -0.073 0" quat="0.707388 0.706825 0 0" mesh="forearm_link_convex_decomposition_p2" name="forearm_link_convex_decomposition_p2_geom" />
|
||||
<body name="wam/wrist_yaw_link" pos="-0.045 0 0" quat="0.707107 0.707107 0 0">
|
||||
<inertial pos="8.921e-05 0.00435824 -0.00511217" quat="0.708528 -0.000120667 0.000107481 0.705683" mass="0.12376" diaginertia="0.0112011 0.0111887 7.58188e-05" />
|
||||
<joint name="wam/wrist_yaw_joint" pos="0 0 0" axis="0 0 1" range="-4.55 1.25" />
|
||||
<geom class="viz" pos="0 0 0.3" mesh="wrist_yaw_link_fine" />
|
||||
<geom class="col" pos="0 0 0.3" mesh="wrist_yaw_link_convex_decomposition_p1" name="wrist_yaw_link_convex_decomposition_p1_geom" />
|
||||
<geom class="col" pos="0 0 0.3" mesh="wrist_yaw_link_convex_decomposition_p2" name="wrist_yaw_link_convex_decomposition_p2_geom" />
|
||||
<body name="wam/wrist_pitch_link" pos="0 0 0.3" quat="0.707107 -0.707107 0 0">
|
||||
<inertial pos="-0.00012262 -0.0246834 -0.0170319" quat="0.994687 -0.102891 0.000824211 -0.00336105" mass="0.417974" diaginertia="0.000555166 0.000463174 0.00023407" />
|
||||
<joint name="wam/wrist_pitch_joint" pos="0 0 0" axis="0 0 1" range="-1.5707 1.5707" />
|
||||
<geom class="viz" mesh="wrist_pitch_link_fine" />
|
||||
<geom class="col" mesh="wrist_pitch_link_convex_decomposition_p1" name="wrist_pitch_link_convex_decomposition_p1_geom" />
|
||||
<geom class="col" mesh="wrist_pitch_link_convex_decomposition_p2" name="wrist_pitch_link_convex_decomposition_p2_geom" />
|
||||
<geom class="col" mesh="wrist_pitch_link_convex_decomposition_p3" name="wrist_pitch_link_convex_decomposition_p3_geom" />
|
||||
<body name="wam/wrist_palm_link" pos="0 -0.06 0" quat="0.707107 0.707107 0 0">
|
||||
<inertial pos="-7.974e-05 -0.00323552 -0.00016313" quat="0.594752 0.382453 0.382453 0.594752" mass="0.0686475" diaginertia="7.408e-05 3.81466e-05 3.76434e-05" />
|
||||
<joint name="wam/palm_yaw_joint" pos="0 0 0" axis="0 0 1" range="-2.7 2.7" />
|
||||
<geom class="viz" pos="0 0 -0.06" mesh="wrist_palm_link_fine" />
|
||||
<geom class="col" pos="0 0 -0.06" mesh="wrist_palm_link_convex" name="wrist_palm_link_convex_geom" />
|
||||
<site name="init_ball_pos_site" size="0.025 0.025 0.025" pos="0.0 0.0 0.035" rgba="0 1 0 1"/>
|
||||
</body>
|
||||
</body>
|
||||
</body>
|
||||
</body>
|
||||
</body>
|
||||
</body>
|
||||
</body>
|
||||
</body>
|
||||
<body name="table_body" pos="0 -1.85 0.4025">
|
||||
<geom name="table" type="box" size="0.4 0.6 0.4" rgba="0.8 0.655 0.45 1" solimp="0.999 0.999 0.001"
|
||||
solref="-10000 -100"/>
|
||||
<geom name="table_contact_geom" type="box" size="0.4 0.6 0.1" pos="0 0 0.31" rgba="1.4 0.8 0.45 1" solimp="0.999 0.999 0.001"
|
||||
solref="-10000 -100"/>
|
||||
</body>
|
||||
<geom name="table_robot" type="box" size="0.1 0.1 0.3" pos="0 0.00 0.3025" rgba="0.8 0.655 0.45 1" solimp="0.999 0.999 0.001"
|
||||
solref="-10000 -100"/>
|
||||
<geom name="wall" type="box" quat="1 0 0 0" size="0.4 0.04 1.1" pos="0. -2.45 1.1" rgba="0.8 0.655 0.45 1" solimp="0.999 0.999 0.001"
|
||||
solref="-10000 -100"/>
|
||||
|
||||
<body name="cup_table" pos="0.32 -1.55 0.84" quat="0.7071068 0.7071068 0 0">
|
||||
<inertial pos="-3.75236e-10 8.27811e-05 0.0947015" quat="0.999945 -0.0104888 0 0" mass="10.132" diaginertia="0.000285643 0.000270485 9.65696e-05" />
|
||||
<geom priority= "1" name="cup_geom_table3" pos="0 0.1 0.001" euler="-1.57 0 0" solref="-10000 -100" type="mesh" mesh="cup3_table" mass="10"/>
|
||||
<geom priority= "1" name="cup_geom_table4" pos="0 0.1 0.001" euler="-1.57 0 0" solref="-10000 -100" type="mesh" mesh="cup4_table" mass="10"/>
|
||||
<geom priority= "1" name="cup_geom_table5" pos="0 0.1 0.001" euler="-1.57 0 0" solref="-10000 -100" type="mesh" mesh="cup5_table" mass="10"/>
|
||||
<geom priority= "1" name="cup_geom_table6" pos="0 0.1 0.001" euler="-1.57 0 0" solref="-10000 -100" type="mesh" mesh="cup6_table" mass="10"/>
|
||||
<geom priority= "1" name="cup_geom_table7" pos="0 0.1 0.001" euler="-1.57 0 0" solref="-10000 -100" type="mesh" mesh="cup7_table" mass="10"/>
|
||||
<geom priority= "1" name="cup_geom_table8" pos="0 0.1 0.001" euler="-1.57 0 0" solref="-10000 -100" type="mesh" mesh="cup8_table" mass="10"/>
|
||||
<geom priority= "1" name="cup_geom_table9" pos="0 0.1 0.001" euler="-1.57 0 0" solref="-10000 -100" type="mesh" mesh="cup9_table" mass="10"/>
|
||||
<geom priority= "1" name="cup_geom_table10" pos="0 0.1 0.001" euler="-1.57 0 0" solref="-10000 -010" type="mesh" mesh="cup10_table" mass="10"/>
|
||||
<geom priority= "1" name="cup_base_table" pos="0 -0.035 0.1337249" euler="-1.57 0 0" type="cylinder" size="0.08 0.045" solref="-10000 -100" mass="10"/>
|
||||
<geom priority= "1" name="cup_base_table_contact" pos="0 0.015 0.1337249" euler="-1.57 0 0" type="cylinder" size="0.07 0.01" solref="-10000 -100" rgba="0 0 255 1" mass="10"/>
|
||||
<geom priority= "1" name="cup_geom_table15" pos="0 0.1 0.001" euler="-1.57 0 0" solref="-10000 -100" type="mesh" mesh="cup15_table" mass="10"/>
|
||||
<geom priority= "1" name="cup_geom_table16" pos="0 0.1 0.001" euler="-1.57 0 0" solref="-10000 -100" type="mesh" mesh="cup16_table" mass="10"/>
|
||||
<geom priority= "1" name="cup_geom_table17" pos="0 0.1 0.001" euler="-1.57 0 0" solref="-10000 -100" type="mesh" mesh="cup17_table" mass="10"/>
|
||||
<geom priority= "1" name="cup_geom1_table8" pos="0 0.1 0.001" euler="-1.57 0 0" solref="-10000 -100" type="mesh" mesh="cup18_table" mass="10"/>
|
||||
<site name="cup_goal_table" pos="0 0.11 0.1337249" rgba="255 0 0 1"/>
|
||||
<site name="cup_goal_final_table" pos="0.0 0.025 0.1337249" rgba="0 255 0 1"/>
|
||||
</body>
|
||||
<body name="ball" pos="0 0 0">
|
||||
<joint axis="1 0 0" damping="0.0" name="tar:x" pos="0 0 0" stiffness="0" type="slide" frictionloss="0" limited="false"/>
|
||||
<joint axis="0 1 0" damping="0.0" name="tar:y" pos="0 0 0" stiffness="0" type="slide" frictionloss="0" limited="false"/>
|
||||
<joint axis="0 0 1" damping="0.0" name="tar:z" pos="0 0 0" stiffness="0" type="slide" frictionloss="0" limited="false"/>
|
||||
<geom priority= "1" size="0.025 0.025 0.025" type="sphere" condim="4" name="ball_geom" rgba="0.8 0.2 0.1 1" mass="0.1"
|
||||
friction="0.1 0.1 0.1" solimp="0.9 0.95 0.001 0.5 2" solref="-10000 -10"/>
|
||||
<site name="target_ball" pos="0 0 0" size="0.02 0.02 0.02" rgba="1 0 0 1" type="sphere"/>
|
||||
</body>
|
||||
<!-- <camera name="visualization" mode="targetbody" target="wam/wrist_yaw_link" pos="1.5 -0.4 2.2"/>-->
|
||||
<!-- <camera name="experiment" mode="fixed" quat="0.44418059 0.41778323 0.54301123 0.57732103" pos="1.5 -0.3 1.33" />-->
|
||||
<camera name="visualization" mode="fixed" euler="2.35 2.0 -0.75" pos="2.0 -0.0 1.85"/>
|
||||
</worldbody>
|
||||
<actuator>
|
||||
<motor ctrllimited="true" ctrlrange="-1.0 1.0" gear="150.0" joint="wam/base_yaw_joint"/>
|
||||
<motor ctrllimited="true" ctrlrange="-1.0 1.0" gear="200.0" joint="wam/shoulder_pitch_joint"/>
|
||||
<motor ctrllimited="true" ctrlrange="-1.0 1.0" gear="50.0" joint="wam/shoulder_yaw_joint"/>
|
||||
<motor ctrllimited="true" ctrlrange="-1.0 1.0" gear="60.0" joint="wam/elbow_pitch_joint"/>
|
||||
<motor ctrllimited="true" ctrlrange="-1.0 1.0" gear="5.0" joint="wam/wrist_yaw_joint"/>
|
||||
<motor ctrllimited="true" ctrlrange="-1.0 1.0" gear="5.0" joint="wam/wrist_pitch_joint"/>
|
||||
<motor ctrllimited="true" ctrlrange="-1.0 1.0" gear="2.0" joint="wam/palm_yaw_joint"/>
|
||||
</actuator>
|
||||
|
||||
</mujoco>
|
||||
@@ -1,3 +1,4 @@
|
||||
import mujoco_py.builder
|
||||
import os
|
||||
|
||||
import numpy as np
|
||||
@@ -5,44 +6,67 @@ from gym import utils
|
||||
from gym.envs.mujoco import MujocoEnv
|
||||
|
||||
|
||||
|
||||
class ALRBeerpongEnv(MujocoEnv, utils.EzPickle):
|
||||
def __init__(self, model_path, frame_skip, n_substeps=4, apply_gravity_comp=True, reward_function=None):
|
||||
utils.EzPickle.__init__(**locals())
|
||||
MujocoEnv.__init__(self, model_path=model_path, frame_skip=frame_skip)
|
||||
class ALRBeerBongEnv(MujocoEnv, utils.EzPickle):
|
||||
def __init__(self, frame_skip=1, apply_gravity_comp=True, reward_type: str = "staged", noisy=False,
|
||||
context: np.ndarray = None, difficulty='simple'):
|
||||
self._steps = 0
|
||||
|
||||
self.xml_path = os.path.join(os.path.dirname(os.path.abspath(__file__)), "assets",
|
||||
"beerpong" + ".xml")
|
||||
|
||||
self.start_pos = np.array([0.0, 1.35, 0.0, 1.18, 0.0, -0.786, -1.59])
|
||||
self.start_vel = np.zeros(7)
|
||||
|
||||
self._q_pos = []
|
||||
self._q_vel = []
|
||||
# self.weight_matrix_scale = 50
|
||||
self.max_ctrl = np.array([150., 125., 40., 60., 5., 5., 2.])
|
||||
self.p_gains = 1 / self.max_ctrl * np.array([200, 300, 100, 100, 10, 10, 2.5])
|
||||
self.d_gains = 1 / self.max_ctrl * np.array([7, 15, 5, 2.5, 0.3, 0.3, 0.05])
|
||||
"beerpong_wo_cup" + ".xml")
|
||||
|
||||
self.j_min = np.array([-2.6, -1.985, -2.8, -0.9, -4.55, -1.5707, -2.7])
|
||||
self.j_max = np.array([2.6, 1.985, 2.8, 3.14159, 1.25, 1.5707, 2.7])
|
||||
|
||||
self.context = None
|
||||
# alr_mujoco_env.AlrMujocoEnv.__init__(self,
|
||||
# self.xml_path,
|
||||
# apply_gravity_comp=apply_gravity_comp,
|
||||
# n_substeps=n_substeps)
|
||||
self.context = context
|
||||
self.apply_gravity_comp = apply_gravity_comp
|
||||
self.add_noise = noisy
|
||||
|
||||
self.sim_time = 8 # seconds
|
||||
self.sim_steps = int(self.sim_time / self.dt)
|
||||
if reward_function is None:
|
||||
from alr_envs.alr.mujoco.beerpong.beerpong_reward import BeerpongReward
|
||||
reward_function = BeerpongReward
|
||||
self.reward_function = reward_function(self.sim, self.sim_steps)
|
||||
self.cup_robot_id = self.sim.model._site_name2id["cup_robot_final"]
|
||||
self.ball_id = self.sim.model._body_name2id["ball"]
|
||||
self.cup_table_id = self.sim.model._body_name2id["cup_table"]
|
||||
self._start_pos = np.array([0.0, 1.35, 0.0, 1.18, 0.0, -0.786, -1.59])
|
||||
self._start_vel = np.zeros(7)
|
||||
|
||||
self.ball_site_id = 0
|
||||
self.ball_id = 11
|
||||
|
||||
self._release_step = 100 # time step of ball release
|
||||
|
||||
self.sim_time = 4 # seconds
|
||||
self.ep_length = 600 # based on 5 seconds with dt = 0.005 int(self.sim_time / self.dt)
|
||||
self.cup_table_id = 10
|
||||
|
||||
if noisy:
|
||||
self.noise_std = 0.01
|
||||
else:
|
||||
self.noise_std = 0
|
||||
|
||||
if difficulty == 'simple':
|
||||
self.cup_goal_pos = np.array([0, -1.7, 0.840])
|
||||
elif difficulty == 'intermediate':
|
||||
self.cup_goal_pos = np.array([0.3, -1.5, 0.840])
|
||||
elif difficulty == 'hard':
|
||||
self.cup_goal_pos = np.array([-0.3, -2.2, 0.840])
|
||||
elif difficulty == 'hardest':
|
||||
self.cup_goal_pos = np.array([-0.3, -1.2, 0.840])
|
||||
|
||||
if reward_type == "no_context":
|
||||
from alr_envs.alr.mujoco.beerpong.beerpong_reward import BeerPongReward
|
||||
reward_function = BeerPongReward
|
||||
elif reward_type == "staged":
|
||||
from alr_envs.alr.mujoco.beerpong.beerpong_reward_staged import BeerPongReward
|
||||
reward_function = BeerPongReward
|
||||
else:
|
||||
raise ValueError("Unknown reward type: {}".format(reward_type))
|
||||
self.reward_function = reward_function()
|
||||
|
||||
MujocoEnv.__init__(self, self.xml_path, frame_skip)
|
||||
utils.EzPickle.__init__(self)
|
||||
|
||||
@property
|
||||
def start_pos(self):
|
||||
return self._start_pos
|
||||
|
||||
@property
|
||||
def start_vel(self):
|
||||
return self._start_vel
|
||||
|
||||
@property
|
||||
def current_pos(self):
|
||||
@@ -52,9 +76,9 @@ class ALRBeerpongEnv(MujocoEnv, utils.EzPickle):
|
||||
def current_vel(self):
|
||||
return self.sim.data.qvel[0:7].copy()
|
||||
|
||||
def configure(self, context):
|
||||
self.context = context
|
||||
self.reward_function.reset(context)
|
||||
def reset(self):
|
||||
self.reward_function.reset(self.add_noise)
|
||||
return super().reset()
|
||||
|
||||
def reset_model(self):
|
||||
init_pos_all = self.init_qpos.copy()
|
||||
@@ -62,19 +86,14 @@ class ALRBeerpongEnv(MujocoEnv, utils.EzPickle):
|
||||
init_vel = np.zeros_like(init_pos_all)
|
||||
|
||||
self._steps = 0
|
||||
self._q_pos = []
|
||||
self._q_vel = []
|
||||
|
||||
start_pos = init_pos_all
|
||||
start_pos[0:7] = init_pos_robot
|
||||
# start_pos[7:] = np.copy(self.sim.data.site_xpos[self.cup_robot_id, :]) + np.array([0., 0.0, 0.05])
|
||||
|
||||
self.set_state(start_pos, init_vel)
|
||||
|
||||
ball_pos = np.copy(self.sim.data.site_xpos[self.cup_robot_id, :]) + np.array([0., 0.0, 0.05])
|
||||
self.sim.model.body_pos[self.ball_id] = ball_pos.copy()
|
||||
self.sim.model.body_pos[self.cup_table_id] = self.context.copy()
|
||||
|
||||
self.sim.model.body_pos[self.cup_table_id] = self.cup_goal_pos
|
||||
start_pos[7::] = self.sim.data.site_xpos[self.ball_site_id, :].copy()
|
||||
self.set_state(start_pos, init_vel)
|
||||
return self._get_obs()
|
||||
|
||||
def step(self, a):
|
||||
@@ -82,32 +101,55 @@ class ALRBeerpongEnv(MujocoEnv, utils.EzPickle):
|
||||
angular_vel = 0.0
|
||||
reward_ctrl = - np.square(a).sum()
|
||||
|
||||
crash = self.do_simulation(a)
|
||||
joint_cons_viol = self.check_traj_in_joint_limits()
|
||||
|
||||
self._q_pos.append(self.sim.data.qpos[0:7].ravel().copy())
|
||||
self._q_vel.append(self.sim.data.qvel[0:7].ravel().copy())
|
||||
if self.apply_gravity_comp:
|
||||
a = a + self.sim.data.qfrc_bias[:len(a)].copy() / self.model.actuator_gear[:, 0]
|
||||
try:
|
||||
self.do_simulation(a, self.frame_skip)
|
||||
if self._steps < self._release_step:
|
||||
self.sim.data.qpos[7::] = self.sim.data.site_xpos[self.ball_site_id, :].copy()
|
||||
self.sim.data.qvel[7::] = self.sim.data.site_xvelp[self.ball_site_id, :].copy()
|
||||
elif self._steps == self._release_step and self.add_noise:
|
||||
self.sim.data.qvel[7::] += self.noise_std * np.random.randn(3)
|
||||
crash = False
|
||||
except mujoco_py.builder.MujocoException:
|
||||
crash = True
|
||||
# joint_cons_viol = self.check_traj_in_joint_limits()
|
||||
|
||||
ob = self._get_obs()
|
||||
|
||||
if not crash and not joint_cons_viol:
|
||||
reward, success, stop_sim = self.reward_function.compute_reward(a, self.sim, self._steps)
|
||||
done = success or self._steps == self.sim_steps - 1 or stop_sim
|
||||
if not crash:
|
||||
reward, reward_infos = self.reward_function.compute_reward(self, a)
|
||||
success = reward_infos['success']
|
||||
is_collided = reward_infos['is_collided']
|
||||
ball_pos = reward_infos['ball_pos']
|
||||
ball_vel = reward_infos['ball_vel']
|
||||
done = is_collided or self._steps == self.ep_length - 1
|
||||
self._steps += 1
|
||||
else:
|
||||
reward = -1000
|
||||
reward = -30
|
||||
success = False
|
||||
is_collided = False
|
||||
done = True
|
||||
ball_pos = np.zeros(3)
|
||||
ball_vel = np.zeros(3)
|
||||
|
||||
return ob, reward, done, dict(reward_dist=reward_dist,
|
||||
reward_ctrl=reward_ctrl,
|
||||
reward=reward,
|
||||
velocity=angular_vel,
|
||||
traj=self._q_pos, is_success=success,
|
||||
is_collided=crash or joint_cons_viol)
|
||||
# traj=self._q_pos,
|
||||
action=a,
|
||||
q_pos=self.sim.data.qpos[0:7].ravel().copy(),
|
||||
q_vel=self.sim.data.qvel[0:7].ravel().copy(),
|
||||
ball_pos=ball_pos,
|
||||
ball_vel=ball_vel,
|
||||
is_success=success,
|
||||
is_collided=is_collided, sim_crash=crash)
|
||||
|
||||
def check_traj_in_joint_limits(self):
|
||||
return any(self.current_pos > self.j_max) or any(self.current_pos < self.j_min)
|
||||
|
||||
# TODO
|
||||
# TODO: extend observation space
|
||||
def _get_obs(self):
|
||||
theta = self.sim.data.qpos.flat[:7]
|
||||
return np.concatenate([
|
||||
@@ -118,28 +160,26 @@ class ALRBeerpongEnv(MujocoEnv, utils.EzPickle):
|
||||
])
|
||||
|
||||
# TODO
|
||||
@property
|
||||
def active_obs(self):
|
||||
pass
|
||||
|
||||
return np.hstack([
|
||||
[False] * 7, # cos
|
||||
[False] * 7, # sin
|
||||
# [True] * 2, # x-y coordinates of target distance
|
||||
[False] # env steps
|
||||
])
|
||||
|
||||
|
||||
if __name__ == "__main__":
|
||||
env = ALRBeerpongEnv()
|
||||
ctxt = np.array([0, -2, 0.840]) # initial
|
||||
env = ALRBeerBongEnv(reward_type="no_context", difficulty='hardest')
|
||||
|
||||
env.configure(ctxt)
|
||||
# env.configure(ctxt)
|
||||
env.reset()
|
||||
env.render()
|
||||
for i in range(16000):
|
||||
# test with random actions
|
||||
ac = 0.0 * env.action_space.sample()[0:7]
|
||||
ac[1] = -0.8
|
||||
ac[3] = - 0.3
|
||||
ac[5] = -0.2
|
||||
# ac = env.start_pos
|
||||
# ac[0] += np.pi/2
|
||||
env.render("human")
|
||||
for i in range(800):
|
||||
ac = 10 * env.action_space.sample()[0:7]
|
||||
obs, rew, d, info = env.step(ac)
|
||||
env.render()
|
||||
env.render("human")
|
||||
|
||||
print(rew)
|
||||
|
||||
@@ -147,4 +187,3 @@ if __name__ == "__main__":
|
||||
break
|
||||
|
||||
env.close()
|
||||
|
||||
|
||||
@@ -1,124 +1,169 @@
|
||||
import numpy as np
|
||||
from alr_envs.alr.mujoco import alr_reward_fct
|
||||
|
||||
|
||||
class BeerpongReward(alr_reward_fct.AlrReward):
|
||||
def __init__(self, sim, sim_time):
|
||||
class BeerPongReward:
|
||||
def __init__(self):
|
||||
|
||||
self.sim = sim
|
||||
self.sim_time = sim_time
|
||||
self.robot_collision_objects = ["wrist_palm_link_convex_geom",
|
||||
"wrist_pitch_link_convex_decomposition_p1_geom",
|
||||
"wrist_pitch_link_convex_decomposition_p2_geom",
|
||||
"wrist_pitch_link_convex_decomposition_p3_geom",
|
||||
"wrist_yaw_link_convex_decomposition_p1_geom",
|
||||
"wrist_yaw_link_convex_decomposition_p2_geom",
|
||||
"forearm_link_convex_decomposition_p1_geom",
|
||||
"forearm_link_convex_decomposition_p2_geom",
|
||||
"upper_arm_link_convex_decomposition_p1_geom",
|
||||
"upper_arm_link_convex_decomposition_p2_geom",
|
||||
"shoulder_link_convex_decomposition_p1_geom",
|
||||
"shoulder_link_convex_decomposition_p2_geom",
|
||||
"shoulder_link_convex_decomposition_p3_geom",
|
||||
"base_link_convex_geom", "table_contact_geom"]
|
||||
|
||||
self.collision_objects = ["cup_geom1", "cup_geom2", "wrist_palm_link_convex_geom",
|
||||
"wrist_pitch_link_convex_decomposition_p1_geom",
|
||||
"wrist_pitch_link_convex_decomposition_p2_geom",
|
||||
"wrist_pitch_link_convex_decomposition_p3_geom",
|
||||
"wrist_yaw_link_convex_decomposition_p1_geom",
|
||||
"wrist_yaw_link_convex_decomposition_p2_geom",
|
||||
"forearm_link_convex_decomposition_p1_geom",
|
||||
"forearm_link_convex_decomposition_p2_geom"]
|
||||
self.cup_collision_objects = ["cup_geom_table3", "cup_geom_table4", "cup_geom_table5", "cup_geom_table6",
|
||||
"cup_geom_table7", "cup_geom_table8", "cup_geom_table9", "cup_geom_table10",
|
||||
# "cup_base_table", "cup_base_table_contact",
|
||||
"cup_geom_table15",
|
||||
"cup_geom_table16",
|
||||
"cup_geom_table17", "cup_geom1_table8",
|
||||
# "cup_base_table_contact",
|
||||
# "cup_base_table"
|
||||
]
|
||||
|
||||
self.ball_id = None
|
||||
self.ball_collision_id = None
|
||||
self.goal_id = None
|
||||
self.goal_final_id = None
|
||||
self.collision_ids = None
|
||||
|
||||
self.ball_traj = None
|
||||
self.dists = None
|
||||
self.dists_ctxt = None
|
||||
self.dists_final = None
|
||||
self.costs = None
|
||||
|
||||
self.action_costs = None
|
||||
self.angle_rewards = None
|
||||
self.cup_angles = None
|
||||
self.cup_z_axes = None
|
||||
self.collision_penalty = 500
|
||||
self.reset(None)
|
||||
|
||||
def reset(self, context):
|
||||
self.ball_traj = np.zeros(shape=(self.sim_time, 3))
|
||||
self.ball_traj = []
|
||||
self.dists = []
|
||||
self.dists_ctxt = []
|
||||
self.dists_final = []
|
||||
self.costs = []
|
||||
self.context = context
|
||||
self.ball_in_cup = False
|
||||
self.dist_ctxt = 5
|
||||
self.action_costs = []
|
||||
self.angle_rewards = []
|
||||
self.cup_angles = []
|
||||
self.cup_z_axes = []
|
||||
self.ball_ground_contact = False
|
||||
self.ball_table_contact = False
|
||||
self.ball_wall_contact = False
|
||||
self.ball_cup_contact = False
|
||||
|
||||
self.ball_id = self.sim.model._body_name2id["ball"]
|
||||
self.ball_collision_id = self.sim.model._geom_name2id["ball_geom"]
|
||||
self.cup_robot_id = self.sim.model._site_name2id["cup_robot_final"]
|
||||
self.goal_id = self.sim.model._site_name2id["cup_goal_table"]
|
||||
self.goal_final_id = self.sim.model._site_name2id["cup_goal_final_table"]
|
||||
self.collision_ids = [self.sim.model._geom_name2id[name] for name in self.collision_objects]
|
||||
self.cup_table_id = self.sim.model._body_name2id["cup_table"]
|
||||
def compute_reward(self, env, action):
|
||||
self.ball_id = env.sim.model._body_name2id["ball"]
|
||||
self.ball_collision_id = env.sim.model._geom_name2id["ball_geom"]
|
||||
self.goal_id = env.sim.model._site_name2id["cup_goal_table"]
|
||||
self.goal_final_id = env.sim.model._site_name2id["cup_goal_final_table"]
|
||||
self.cup_collision_ids = [env.sim.model._geom_name2id[name] for name in self.cup_collision_objects]
|
||||
self.cup_table_id = env.sim.model._body_name2id["cup_table"]
|
||||
self.table_collision_id = env.sim.model._geom_name2id["table_contact_geom"]
|
||||
self.wall_collision_id = env.sim.model._geom_name2id["wall"]
|
||||
self.cup_table_collision_id = env.sim.model._geom_name2id["cup_base_table_contact"]
|
||||
self.init_ball_pos_site_id = env.sim.model._site_name2id["init_ball_pos_site"]
|
||||
self.ground_collision_id = env.sim.model._geom_name2id["ground"]
|
||||
self.robot_collision_ids = [env.sim.model._geom_name2id[name] for name in self.robot_collision_objects]
|
||||
|
||||
def compute_reward(self, action, sim, step):
|
||||
action_cost = np.sum(np.square(action))
|
||||
|
||||
stop_sim = False
|
||||
success = False
|
||||
|
||||
if self.check_collision(sim):
|
||||
reward = - 1e-4 * action_cost - 1000
|
||||
stop_sim = True
|
||||
return reward, success, stop_sim
|
||||
|
||||
# Compute the current distance from the ball to the inner part of the cup
|
||||
goal_pos = sim.data.site_xpos[self.goal_id]
|
||||
ball_pos = sim.data.body_xpos[self.ball_id]
|
||||
goal_final_pos = sim.data.site_xpos[self.goal_final_id]
|
||||
goal_pos = env.sim.data.site_xpos[self.goal_id]
|
||||
ball_pos = env.sim.data.body_xpos[self.ball_id]
|
||||
goal_final_pos = env.sim.data.site_xpos[self.goal_final_id]
|
||||
self.dists.append(np.linalg.norm(goal_pos - ball_pos))
|
||||
self.dists_final.append(np.linalg.norm(goal_final_pos - ball_pos))
|
||||
self.dists_ctxt.append(np.linalg.norm(ball_pos - self.context))
|
||||
self.ball_traj[step, :] = ball_pos
|
||||
|
||||
# Determine the first time when ball is in cup
|
||||
if not self.ball_in_cup:
|
||||
ball_in_cup = self.check_ball_in_cup(sim, self.ball_collision_id)
|
||||
self.ball_in_cup = ball_in_cup
|
||||
if ball_in_cup:
|
||||
dist_to_ctxt = np.linalg.norm(ball_pos - self.context)
|
||||
self.dist_ctxt = dist_to_ctxt
|
||||
action_cost = np.sum(np.square(action))
|
||||
self.action_costs.append(action_cost)
|
||||
|
||||
ball_table_bounce = self._check_collision_single_objects(env.sim, self.ball_collision_id,
|
||||
self.table_collision_id)
|
||||
|
||||
if ball_table_bounce: # or ball_cup_table_cont or ball_wall_con
|
||||
self.ball_table_contact = True
|
||||
|
||||
ball_cup_cont = self._check_collision_with_set_of_objects(env.sim, self.ball_collision_id,
|
||||
self.cup_collision_ids)
|
||||
if ball_cup_cont:
|
||||
self.ball_cup_contact = True
|
||||
|
||||
ball_wall_cont = self._check_collision_single_objects(env.sim, self.ball_collision_id, self.wall_collision_id)
|
||||
if ball_wall_cont and not self.ball_table_contact:
|
||||
self.ball_wall_contact = True
|
||||
|
||||
ball_ground_contact = self._check_collision_single_objects(env.sim, self.ball_collision_id,
|
||||
self.ground_collision_id)
|
||||
if ball_ground_contact and not self.ball_table_contact:
|
||||
self.ball_ground_contact = True
|
||||
|
||||
self._is_collided = self._check_collision_with_itself(env.sim, self.robot_collision_ids)
|
||||
if env._steps == env.ep_length - 1 or self._is_collided:
|
||||
|
||||
if step == self.sim_time - 1:
|
||||
min_dist = np.min(self.dists)
|
||||
dist_final = self.dists_final[-1]
|
||||
# dist_ctxt = self.dists_ctxt[-1]
|
||||
|
||||
# cost = self._get_stage_wise_cost(ball_in_cup, min_dist, dist_final, dist_ctxt)
|
||||
cost = 2 * (0.5 * min_dist + 0.5 * dist_final + 0.1 * self.dist_ctxt)
|
||||
reward = np.exp(-1 * cost) - 1e-4 * action_cost
|
||||
success = dist_final < 0.05 and self.dist_ctxt < 0.05
|
||||
ball_in_cup = self._check_collision_single_objects(env.sim, self.ball_collision_id, self.cup_table_collision_id)
|
||||
|
||||
cost_offset = 0
|
||||
|
||||
if self.ball_ground_contact: # or self.ball_wall_contact:
|
||||
cost_offset += 2
|
||||
|
||||
if not self.ball_table_contact:
|
||||
cost_offset += 2
|
||||
|
||||
if not ball_in_cup:
|
||||
cost_offset += 2
|
||||
cost = cost_offset + min_dist ** 2 + 0.5 * self.dists_final[-1] ** 2 + 1e-4 * action_cost # + min_dist ** 2
|
||||
else:
|
||||
if self.ball_cup_contact:
|
||||
cost_offset += 1
|
||||
cost = cost_offset + self.dists_final[-1] ** 2 + 1e-4 * action_cost
|
||||
|
||||
reward = - 1*cost - self.collision_penalty * int(self._is_collided)
|
||||
success = ball_in_cup and not self.ball_ground_contact and not self.ball_wall_contact and not self.ball_cup_contact
|
||||
else:
|
||||
reward = - 1e-4 * action_cost
|
||||
success = False
|
||||
|
||||
return reward, success, stop_sim
|
||||
infos = {}
|
||||
infos["success"] = success
|
||||
infos["is_collided"] = self._is_collided
|
||||
infos["ball_pos"] = ball_pos.copy()
|
||||
infos["action_cost"] = 5e-4 * action_cost
|
||||
|
||||
def _get_stage_wise_cost(self, ball_in_cup, min_dist, dist_final, dist_to_ctxt):
|
||||
if not ball_in_cup:
|
||||
cost = 3 + 2*(0.5 * min_dist**2 + 0.5 * dist_final**2)
|
||||
else:
|
||||
cost = 2 * dist_to_ctxt ** 2
|
||||
print('Context Distance:', dist_to_ctxt)
|
||||
return cost
|
||||
return reward, infos
|
||||
|
||||
def check_ball_in_cup(self, sim, ball_collision_id):
|
||||
cup_base_collision_id = sim.model._geom_name2id["cup_base_contact"]
|
||||
def _check_collision_single_objects(self, sim, id_1, id_2):
|
||||
for coni in range(0, sim.data.ncon):
|
||||
con = sim.data.contact[coni]
|
||||
|
||||
collision = con.geom1 == cup_base_collision_id and con.geom2 == ball_collision_id
|
||||
collision_trans = con.geom1 == ball_collision_id and con.geom2 == cup_base_collision_id
|
||||
collision = con.geom1 == id_1 and con.geom2 == id_2
|
||||
collision_trans = con.geom1 == id_2 and con.geom2 == id_1
|
||||
|
||||
if collision or collision_trans:
|
||||
return True
|
||||
return False
|
||||
|
||||
def check_collision(self, sim):
|
||||
def _check_collision_with_itself(self, sim, collision_ids):
|
||||
col_1, col_2 = False, False
|
||||
for j, id in enumerate(collision_ids):
|
||||
col_1 = self._check_collision_with_set_of_objects(sim, id, collision_ids[:j])
|
||||
if j != len(collision_ids) - 1:
|
||||
col_2 = self._check_collision_with_set_of_objects(sim, id, collision_ids[j + 1:])
|
||||
else:
|
||||
col_2 = False
|
||||
collision = True if col_1 or col_2 else False
|
||||
return collision
|
||||
|
||||
def _check_collision_with_set_of_objects(self, sim, id_1, id_list):
|
||||
for coni in range(0, sim.data.ncon):
|
||||
con = sim.data.contact[coni]
|
||||
|
||||
collision = con.geom1 in self.collision_ids and con.geom2 == self.ball_collision_id
|
||||
collision_trans = con.geom1 == self.ball_collision_id and con.geom2 in self.collision_ids
|
||||
collision = con.geom1 in id_list and con.geom2 == id_1
|
||||
collision_trans = con.geom1 == id_1 and con.geom2 in id_list
|
||||
|
||||
if collision or collision_trans:
|
||||
return True
|
||||
return False
|
||||
return False
|
||||
@@ -0,0 +1,169 @@
|
||||
import numpy as np
|
||||
|
||||
|
||||
class BeerPongReward:
|
||||
def __init__(self):
|
||||
|
||||
self.robot_collision_objects = ["wrist_palm_link_convex_geom",
|
||||
"wrist_pitch_link_convex_decomposition_p1_geom",
|
||||
"wrist_pitch_link_convex_decomposition_p2_geom",
|
||||
"wrist_pitch_link_convex_decomposition_p3_geom",
|
||||
"wrist_yaw_link_convex_decomposition_p1_geom",
|
||||
"wrist_yaw_link_convex_decomposition_p2_geom",
|
||||
"forearm_link_convex_decomposition_p1_geom",
|
||||
"forearm_link_convex_decomposition_p2_geom",
|
||||
"upper_arm_link_convex_decomposition_p1_geom",
|
||||
"upper_arm_link_convex_decomposition_p2_geom",
|
||||
"shoulder_link_convex_decomposition_p1_geom",
|
||||
"shoulder_link_convex_decomposition_p2_geom",
|
||||
"shoulder_link_convex_decomposition_p3_geom",
|
||||
"base_link_convex_geom", "table_contact_geom"]
|
||||
|
||||
self.cup_collision_objects = ["cup_geom_table3", "cup_geom_table4", "cup_geom_table5", "cup_geom_table6",
|
||||
"cup_geom_table7", "cup_geom_table8", "cup_geom_table9", "cup_geom_table10",
|
||||
# "cup_base_table", "cup_base_table_contact",
|
||||
"cup_geom_table15",
|
||||
"cup_geom_table16",
|
||||
"cup_geom_table17", "cup_geom1_table8",
|
||||
# "cup_base_table_contact",
|
||||
# "cup_base_table"
|
||||
]
|
||||
|
||||
|
||||
self.ball_traj = None
|
||||
self.dists = None
|
||||
self.dists_final = None
|
||||
self.costs = None
|
||||
self.action_costs = None
|
||||
self.angle_rewards = None
|
||||
self.cup_angles = None
|
||||
self.cup_z_axes = None
|
||||
self.collision_penalty = 500
|
||||
self.reset(None)
|
||||
|
||||
def reset(self, noisy):
|
||||
self.ball_traj = []
|
||||
self.dists = []
|
||||
self.dists_final = []
|
||||
self.costs = []
|
||||
self.action_costs = []
|
||||
self.angle_rewards = []
|
||||
self.cup_angles = []
|
||||
self.cup_z_axes = []
|
||||
self.ball_ground_contact = False
|
||||
self.ball_table_contact = False
|
||||
self.ball_wall_contact = False
|
||||
self.ball_cup_contact = False
|
||||
self.noisy_bp = noisy
|
||||
self._t_min_final_dist = -1
|
||||
|
||||
def compute_reward(self, env, action):
|
||||
self.ball_id = env.sim.model._body_name2id["ball"]
|
||||
self.ball_collision_id = env.sim.model._geom_name2id["ball_geom"]
|
||||
self.goal_id = env.sim.model._site_name2id["cup_goal_table"]
|
||||
self.goal_final_id = env.sim.model._site_name2id["cup_goal_final_table"]
|
||||
self.cup_collision_ids = [env.sim.model._geom_name2id[name] for name in self.cup_collision_objects]
|
||||
self.cup_table_id = env.sim.model._body_name2id["cup_table"]
|
||||
self.table_collision_id = env.sim.model._geom_name2id["table_contact_geom"]
|
||||
self.wall_collision_id = env.sim.model._geom_name2id["wall"]
|
||||
self.cup_table_collision_id = env.sim.model._geom_name2id["cup_base_table_contact"]
|
||||
self.init_ball_pos_site_id = env.sim.model._site_name2id["init_ball_pos_site"]
|
||||
self.ground_collision_id = env.sim.model._geom_name2id["ground"]
|
||||
self.robot_collision_ids = [env.sim.model._geom_name2id[name] for name in self.robot_collision_objects]
|
||||
|
||||
goal_pos = env.sim.data.site_xpos[self.goal_id]
|
||||
ball_pos = env.sim.data.body_xpos[self.ball_id]
|
||||
ball_vel = env.sim.data.body_xvelp[self.ball_id]
|
||||
goal_final_pos = env.sim.data.site_xpos[self.goal_final_id]
|
||||
self.dists.append(np.linalg.norm(goal_pos - ball_pos))
|
||||
self.dists_final.append(np.linalg.norm(goal_final_pos - ball_pos))
|
||||
|
||||
action_cost = np.sum(np.square(action))
|
||||
self.action_costs.append(action_cost)
|
||||
|
||||
self._is_collided = self._check_collision_with_itself(env.sim, self.robot_collision_ids)
|
||||
if env._steps == env.ep_length - 1 or self._is_collided:
|
||||
|
||||
min_dist = np.min(self.dists)
|
||||
ball_table_bounce = self._check_collision_single_objects(env.sim, self.ball_collision_id,
|
||||
self.table_collision_id)
|
||||
ball_cup_table_cont = self._check_collision_with_set_of_objects(env.sim, self.ball_collision_id,
|
||||
self.cup_collision_ids)
|
||||
ball_wall_cont = self._check_collision_single_objects(env.sim, self.ball_collision_id,
|
||||
self.wall_collision_id)
|
||||
ball_in_cup = self._check_collision_single_objects(env.sim, self.ball_collision_id,
|
||||
self.cup_table_collision_id)
|
||||
if not ball_in_cup:
|
||||
cost_offset = 2
|
||||
if not ball_cup_table_cont and not ball_table_bounce and not ball_wall_cont:
|
||||
cost_offset += 2
|
||||
cost = cost_offset + min_dist ** 2 + 0.5 * self.dists_final[-1] ** 2 + 1e-7 * action_cost
|
||||
else:
|
||||
cost = self.dists_final[-1] ** 2 + 1.5 * action_cost * 1e-7
|
||||
|
||||
reward = - 1 * cost - self.collision_penalty * int(self._is_collided)
|
||||
success = ball_in_cup
|
||||
crash = self._is_collided
|
||||
else:
|
||||
reward = - 1e-7 * action_cost
|
||||
cost = 0
|
||||
success = False
|
||||
crash = False
|
||||
|
||||
infos = {}
|
||||
infos["success"] = success
|
||||
infos["is_collided"] = self._is_collided
|
||||
infos["ball_pos"] = ball_pos.copy()
|
||||
infos["ball_vel"] = ball_vel.copy()
|
||||
infos["action_cost"] = 5e-4 * action_cost
|
||||
infos["task_cost"] = cost
|
||||
|
||||
return reward, infos
|
||||
|
||||
def get_cost_offset(self):
|
||||
if self.ball_ground_contact:
|
||||
return 200
|
||||
|
||||
if not self.ball_table_contact:
|
||||
return 100
|
||||
|
||||
if not self.ball_in_cup:
|
||||
return 50
|
||||
|
||||
if self.ball_in_cup and self.ball_cup_contact and not self.noisy_bp:
|
||||
return 10
|
||||
|
||||
return 0
|
||||
|
||||
def _check_collision_single_objects(self, sim, id_1, id_2):
|
||||
for coni in range(0, sim.data.ncon):
|
||||
con = sim.data.contact[coni]
|
||||
|
||||
collision = con.geom1 == id_1 and con.geom2 == id_2
|
||||
collision_trans = con.geom1 == id_2 and con.geom2 == id_1
|
||||
|
||||
if collision or collision_trans:
|
||||
return True
|
||||
return False
|
||||
|
||||
def _check_collision_with_itself(self, sim, collision_ids):
|
||||
col_1, col_2 = False, False
|
||||
for j, id in enumerate(collision_ids):
|
||||
col_1 = self._check_collision_with_set_of_objects(sim, id, collision_ids[:j])
|
||||
if j != len(collision_ids) - 1:
|
||||
col_2 = self._check_collision_with_set_of_objects(sim, id, collision_ids[j + 1:])
|
||||
else:
|
||||
col_2 = False
|
||||
collision = True if col_1 or col_2 else False
|
||||
return collision
|
||||
|
||||
def _check_collision_with_set_of_objects(self, sim, id_1, id_list):
|
||||
for coni in range(0, sim.data.ncon):
|
||||
con = sim.data.contact[coni]
|
||||
|
||||
collision = con.geom1 in id_list and con.geom2 == id_1
|
||||
collision_trans = con.geom1 == id_1 and con.geom2 in id_list
|
||||
|
||||
if collision or collision_trans:
|
||||
return True
|
||||
return False
|
||||
@@ -0,0 +1,39 @@
|
||||
from typing import Tuple, Union
|
||||
|
||||
import numpy as np
|
||||
|
||||
from mp_env_api.interface_wrappers.mp_env_wrapper import MPEnvWrapper
|
||||
|
||||
|
||||
class MPWrapper(MPEnvWrapper):
|
||||
|
||||
@property
|
||||
def active_obs(self):
|
||||
# TODO: @Max Filter observations correctly
|
||||
return np.hstack([
|
||||
[False] * 7, # cos
|
||||
[False] * 7, # sin
|
||||
# [True] * 2, # x-y coordinates of target distance
|
||||
[False] # env steps
|
||||
])
|
||||
|
||||
@property
|
||||
def start_pos(self):
|
||||
return self._start_pos
|
||||
|
||||
@property
|
||||
def current_pos(self) -> Union[float, int, np.ndarray, Tuple]:
|
||||
return self.sim.data.qpos[0:7].copy()
|
||||
|
||||
@property
|
||||
def current_vel(self) -> Union[float, int, np.ndarray, Tuple]:
|
||||
return self.sim.data.qvel[0:7].copy()
|
||||
|
||||
@property
|
||||
def goal_pos(self):
|
||||
# TODO: @Max I think the default value of returning to the start is reasonable here
|
||||
raise ValueError("Goal position is not available and has to be learnt based on the environment.")
|
||||
|
||||
@property
|
||||
def dt(self) -> Union[float, int]:
|
||||
return self.env.dt
|
||||
Reference in New Issue
Block a user