File size: 10,227 Bytes
7985e7e
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
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
169
170
171
172
173
174
175
176
177
178
179
180
181
182
183
184
185
186
187
188
189
190
191
192
193
194
195
196
197
198
199
200
201
202
203
204
205
206
207
208
209
210
211
212
213
214
215
216
217
218
219
220
221
222
223
224
225
226
227
228
229
230
231
232
233
234
235
236
237
238
239
240
241
242
243
244
245
246
247
248
249
250
251
252
253
254
255
256
257
258
259
260
261
262
263
264
265
266
267
268
269
270
271
272
273
274
275
"""
DiReCTPolicy: Diffusion with Real-time Constrained Trajectory optimization.

At each denoising step, predicts x_0 via the diffusion model, projects
the clean prediction through IPOPT (obstacle + dynamics + action constraints),
then maps the correction back to noisy space.
"""

import math
import time
from collections import namedtuple

import numpy as np
import torch
from torch import nn

from model import CosineScheduleDiffusion, apply_conditioning
from obstacles import CONSTRAINT_SETS
from utils.arrays import to_np, to_torch, apply_dict

Trajectories = namedtuple("Trajectories", "actions observations values")


class DiReCTPolicy(nn.Module):
    """DDPM sampling with IPOPT projection in clean x_0 space.

    At each denoising step where guidance is active, the model's x_0 prediction
    is projected onto the feasible set via IPOPT, and the correction is mapped
    back to noisy space via alpha_s * (x0_projected - x0_pred).
    """

    def __init__(
        self,
        model,
        normalizer,
        action_dim,
        state_dim,
        horizon,
        n_sampling_steps=10,
        guidance_start=0.0,
        guidance_last_steps=0,
        obstacle_margin=0.02,
        control_penalty_weight=1.0,
        prediction_objective="",
        prediction_objective_scale=100.0,
        max_action_delta=0.0,
        dynamics_model=None,
        dynamics_relaxation=0.0,
        device="cuda:0",
    ):
        super().__init__()
        self.model = model
        self.normalizer = normalizer
        self.action_dim = action_dim
        self.state_dim = state_dim
        self.horizon = horizon
        self.n_sampling_steps = n_sampling_steps
        self.guidance_start = guidance_start
        self.guidance_last_steps = guidance_last_steps
        self.obstacle_margin = obstacle_margin
        self.control_penalty_weight = control_penalty_weight
        self.prediction_objective = prediction_objective
        self.prediction_objective_scale = prediction_objective_scale
        self.max_action_delta = max_action_delta
        self.device = device

        self.diffusion = CosineScheduleDiffusion(model, action_dim)

        # Build IPOPT projector
        from projector import IpoptProjector

        circular_obs = CONSTRAINT_SETS["novel"]["circular"]
        planar_obs = CONSTRAINT_SETS["novel"]["planar"]

        objective_fn = None
        if prediction_objective == "distance":
            objective_fn = self._make_distance_objective(normalizer)

        self._projector = IpoptProjector(
            circular_obstacles=circular_obs,
            planar_obstacles=planar_obs,
            horizon=horizon,
            action_dim=action_dim,
            state_dim=state_dim,
            normalizer=normalizer,
            dynamics=dynamics_model,
            dynamics_relaxation=dynamics_relaxation,
            obstacle_margin=obstacle_margin,
            objective_fn=objective_fn,
            objective_scale=prediction_objective_scale,
            max_action_delta=max_action_delta,
        )

    def _make_distance_objective(self, normalizer):
        """Create CasADi objective: minimize distance to goal y-position."""
        import casadi as cs

        obs_mins = normalizer.normalizers["observations"].mins
        obs_maxs = normalizer.normalizers["observations"].maxs
        ymin = float(obs_mins[3])
        ymax = float(obs_maxs[3])
        y_goal = 0.35
        dof = self.horizon * (self.action_dim + self.state_dim) - self.state_dim

        def obj_fn(X_dof, s0):
            y_norm = X_dof[dof - self.state_dim + 3]
            y_real = (y_norm + 1.0) / 2.0 * (ymax - ymin) + ymin
            return (y_real - y_goal) ** 2
        return obj_fn

    def __call__(self, conditions, batch_size=1, a_prev=None):
        """Sample trajectories with IPOPT-projected DDPM reverse process."""
        conditions = apply_dict(
            self.normalizer.normalize, conditions, "observations"
        )

        t_start = time.time()

        transition_dim = self.action_dim + self.state_dim
        conditions_torch = to_torch(conditions, device=self.device)

        x = torch.randn(
            batch_size, self.horizon, transition_dim, device=self.device
        )
        x = apply_conditioning(x, conditions_torch, self.action_dim)

        # per-element IPOPT statistics accumulated over the denoise loop
        self._ipopt_solves = 0
        self._ipopt_iters = np.zeros(batch_size)
        self._ipopt_failures = np.zeros(batch_size)

        timesteps = torch.linspace(
            1.0, 0.0, self.n_sampling_steps + 1, device=self.device
        )

        save_chain = batch_size == 1
        if save_chain:
            denoising_chain = [x.clone()]
            x0_estimates = []

        for i in range(self.n_sampling_steps):
            t_now = timesteps[i]
            t_next = timesteps[i + 1]
            is_last_step = i == self.n_sampling_steps - 1
            dt = t_now - t_next

            abar_now = self.diffusion.alpha_bar(t_now)
            abar_next = self.diffusion.alpha_bar(t_next)
            g_t = self.diffusion.diffusion_coeff(t_now)

            t_batch = t_now.expand(batch_size)
            x0_pred = self.model(x, t_batch)
            x0_pred = apply_conditioning(x0_pred, conditions_torch, self.action_dim)

            if save_chain:
                x0_estimates.append(x0_pred.detach().clone())

            beta = 1.0 - abar_now / abar_next
            coeff_x0 = math.sqrt(abar_next) * beta / (1.0 - abar_now)
            coeff_xt = math.sqrt(1.0 - beta) * (1.0 - abar_next) / (1.0 - abar_now)
            posterior_mean = coeff_x0 * x0_pred + coeff_xt * x
            posterior_var = max((1.0 - abar_next) / (1.0 - abar_now) * beta, 0.0)

            if is_last_step:
                x_uncontrolled = posterior_mean
            else:
                noise = torch.randn_like(x)
                x_uncontrolled = posterior_mean + math.sqrt(posterior_var) * noise

            x_uncontrolled = apply_conditioning(
                x_uncontrolled, conditions_torch, self.action_dim
            )

            frac_done = i / self.n_sampling_steps
            if self.guidance_last_steps > 0:
                steps_remaining = self.n_sampling_steps - i
                apply_guidance = steps_remaining <= self.guidance_last_steps
            else:
                apply_guidance = frac_done >= self.guidance_start

            if apply_guidance:
                x_next = self._apply_prediction_guidance(
                    x, x0_pred, x_uncontrolled, dt,
                    abar_next, conditions_torch, batch_size,
                    a_prev=a_prev,
                    g_t=g_t,
                )
            else:
                x_next = x_uncontrolled

            x = apply_conditioning(x_next, conditions_torch, self.action_dim)
            x = x.detach()

            if save_chain:
                denoising_chain.append(x.clone())

        normed_actions = x[:, :, : self.action_dim]
        actions = self.normalizer.unnormalize(to_np(normed_actions), "actions")

        normed_observations = x[:, :, self.action_dim :]
        observations = self.normalizer.unnormalize(
            to_np(normed_observations), "observations"
        )

        trajectories = Trajectories(actions, observations, np.zeros((batch_size, 1)))

        t_end = time.time()
        info = {"computation_time": t_end - t_start}
        if self._ipopt_solves > 0:
            info["ipopt_solves"] = np.full(batch_size, float(self._ipopt_solves))
            info["ipopt_iters"] = self._ipopt_iters
            info["ipopt_failures"] = self._ipopt_failures

        if save_chain:
            chain = torch.stack(denoising_chain, dim=1)
            chain = self._unnormalize_chain(chain)
            x0_est = torch.stack(x0_estimates, dim=1)
            x0_est = self._unnormalize_chain(x0_est)
            return actions[:, 0], trajectories, chain, x0_est, info
        else:
            return actions[:, 0], trajectories, None, None, info

    def _apply_prediction_guidance(
        self, x, x0_pred, x_uncontrolled, dt, abar_next,
        conditions_torch, batch_size, a_prev=None, g_t=None,
    ):
        """Project x0_pred in clean space via IPOPT, map correction back to noisy space."""
        B = batch_size
        x0_np = to_np(x0_pred)

        dofs = np.stack([
            self._projector.full_trajectory_to_dof(x0_np[b]) for b in range(B)
        ])
        s0s = x0_np[:, 0, self.action_dim:]

        # Time-dependent regularization: lambda * alpha_s^2 / dt
        # reg_weight = float(self.control_penalty_weight * float(abar_next) / float(dt))

        ## with g**2 scaling
        reg_weight = float(self.control_penalty_weight * float(abar_next) / float(dt) / float(g_t)**2)

        a_prevs = None
        if a_prev is not None and self.max_action_delta > 0:
            a_prevs = np.tile(to_np(a_prev).reshape(1, -1), (B, 1))

        dofs_proj, statuses, iters = self._projector.project_batch(
            dofs, s0s, n_workers=B, reg_weight=reg_weight, a_prevs=a_prevs
        )
        self._ipopt_solves += 1
        self._ipopt_iters += np.asarray(iters, dtype=float)
        self._ipopt_failures += np.asarray([s == "failed" for s in statuses], dtype=float)

        x0_proj_np = np.stack([
            self._projector.dof_to_full_trajectory(dofs_proj[b], s0s[b])
            for b in range(B)
        ])
        x0_projected = torch.tensor(
            x0_proj_np, dtype=x.dtype, device=x.device
        )
        x0_projected = apply_conditioning(
            x0_projected, conditions_torch, self.action_dim
        )

        alpha_s = math.sqrt(float(abar_next))
        x_next = x_uncontrolled + alpha_s * (x0_projected - x0_pred)
        return x_next

    def _unnormalize_chain(self, chain):
        """Unnormalize a chain of shape (batch, steps, horizon, transition_dim)."""
        obs_chain = chain[..., self.action_dim :]
        obs_unnorm = self.normalizer.unnormalize(to_np(obs_chain), "observations")
        act_chain = chain[..., : self.action_dim]
        act_unnorm = self.normalizer.unnormalize(to_np(act_chain), "actions")
        return np.concatenate([act_unnorm, obs_unnorm], axis=-1)