File size: 4,942 Bytes
375a1cf
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
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
151
152
153
154
155
156
157
158
159
160
161
162
163
164
165
166
167
168
import numpy as np

from gym import utils
from gym.envs.mujoco import MuJocoPyEnv
from gym.spaces import Box

DEFAULT_CAMERA_CONFIG = {
    "trackbodyid": 2,
    "distance": 4.0,
    "lookat": np.array((0.0, 0.0, 1.15)),
    "elevation": -20.0,
}


class Walker2dEnv(MuJocoPyEnv, utils.EzPickle):
    metadata = {
        "render_modes": [
            "human",
            "rgb_array",
            "depth_array",
        ],
        "render_fps": 125,
    }

    def __init__(
        self,
        xml_file="walker2d.xml",
        forward_reward_weight=1.0,
        ctrl_cost_weight=1e-3,
        healthy_reward=1.0,
        terminate_when_unhealthy=True,
        healthy_z_range=(0.8, 2.0),
        healthy_angle_range=(-1.0, 1.0),
        reset_noise_scale=5e-3,
        exclude_current_positions_from_observation=True,
        **kwargs
    ):
        utils.EzPickle.__init__(
            self,
            xml_file,
            forward_reward_weight,
            ctrl_cost_weight,
            healthy_reward,
            terminate_when_unhealthy,
            healthy_z_range,
            healthy_angle_range,
            reset_noise_scale,
            exclude_current_positions_from_observation,
            **kwargs
        )

        self._forward_reward_weight = forward_reward_weight
        self._ctrl_cost_weight = ctrl_cost_weight

        self._healthy_reward = healthy_reward
        self._terminate_when_unhealthy = terminate_when_unhealthy

        self._healthy_z_range = healthy_z_range
        self._healthy_angle_range = healthy_angle_range

        self._reset_noise_scale = reset_noise_scale

        self._exclude_current_positions_from_observation = (
            exclude_current_positions_from_observation
        )

        if exclude_current_positions_from_observation:
            observation_space = Box(
                low=-np.inf, high=np.inf, shape=(17,), dtype=np.float64
            )
        else:
            observation_space = Box(
                low=-np.inf, high=np.inf, shape=(18,), dtype=np.float64
            )

        MuJocoPyEnv.__init__(
            self, xml_file, 4, observation_space=observation_space, **kwargs
        )

    @property
    def healthy_reward(self):
        return (
            float(self.is_healthy or self._terminate_when_unhealthy)
            * self._healthy_reward
        )

    def control_cost(self, action):
        control_cost = self._ctrl_cost_weight * np.sum(np.square(action))
        return control_cost

    @property
    def is_healthy(self):
        z, angle = self.sim.data.qpos[1:3]

        min_z, max_z = self._healthy_z_range
        min_angle, max_angle = self._healthy_angle_range

        healthy_z = min_z < z < max_z
        healthy_angle = min_angle < angle < max_angle
        is_healthy = healthy_z and healthy_angle

        return is_healthy

    @property
    def terminated(self):
        terminated = not self.is_healthy if self._terminate_when_unhealthy else False
        return terminated

    def _get_obs(self):
        position = self.sim.data.qpos.flat.copy()
        velocity = np.clip(self.sim.data.qvel.flat.copy(), -10, 10)

        if self._exclude_current_positions_from_observation:
            position = position[1:]

        observation = np.concatenate((position, velocity)).ravel()
        return observation

    def step(self, action):
        x_position_before = self.sim.data.qpos[0]
        self.do_simulation(action, self.frame_skip)
        x_position_after = self.sim.data.qpos[0]
        x_velocity = (x_position_after - x_position_before) / self.dt

        ctrl_cost = self.control_cost(action)
        forward_reward = self._forward_reward_weight * x_velocity
        healthy_reward = self.healthy_reward

        rewards = forward_reward + healthy_reward
        costs = ctrl_cost

        observation = self._get_obs()
        reward = rewards - costs
        terminated = self.terminated
        info = {
            "x_position": x_position_after,
            "x_velocity": x_velocity,
        }

        if self.render_mode == "human":
            self.render()

        return observation, reward, terminated, False, info

    def reset_model(self):
        noise_low = -self._reset_noise_scale
        noise_high = self._reset_noise_scale

        qpos = self.init_qpos + self.np_random.uniform(
            low=noise_low, high=noise_high, size=self.model.nq
        )
        qvel = self.init_qvel + self.np_random.uniform(
            low=noise_low, high=noise_high, size=self.model.nv
        )

        self.set_state(qpos, qvel)

        observation = self._get_obs()
        return observation

    def viewer_setup(self):
        assert self.viewer is not None
        for key, value in DEFAULT_CAMERA_CONFIG.items():
            if isinstance(value, np.ndarray):
                getattr(self.viewer.cam, key)[:] = value
            else:
                setattr(self.viewer.cam, key, value)