openpilot_comma/selfdrive/locationd/lagd.py

#!/usr/bin/env python3
import numpy as np
from collections import deque
from functools import partial

import cereal.messaging as messaging
from cereal import car, log
from cereal.services import SERVICE_LIST
from openpilot.common.params import Params
from openpilot.common.realtime import config_realtime_process, DT_CTRL
from openpilot.selfdrive.locationd.helpers import PoseCalibrator, Pose

MIN_LAG_VEL = 20.0
MAX_SANE_LAG = 3.0
MIN_ABS_YAW_RATE_DEG = 1
MOVING_CORR_WINDOW = 300.0
MIN_OKAY_WINDOW = 25.0
MIN_NCC = 0.95


class Points:
  def __init__(self, num_points):
    self.times = deque(maxlen=num_points)
    self.okay = deque(maxlen=num_points)
    self.desired = deque(maxlen=num_points)
    self.actual = deque(maxlen=num_points)

  @property
  def num_points(self):
    return len(self.desired)

  @property
  def num_okay(self):
    return np.count_nonzero(self.okay)

  def update(self, t, desired, actual, okay):
    self.times.append(t)
    self.okay.append(okay)
    self.desired.append(desired)
    self.actual.append(actual)

  def get(self):
    return np.array(self.times), np.array(self.desired), np.array(self.actual), np.array(self.okay)


class BlockAverage:
  def __init__(self, num_blocks, block_size, valid_blocks, initial_value):
    self.num_blocks = num_blocks
    self.block_size = block_size
    self.block_idx = 0
    self.idx = 0

    self.values = np.tile(initial_value, (num_blocks, 1))
    self.valid_blocks = valid_blocks

  def update(self, value):
    self.values[self.block_idx] = (self.idx * self.values[self.block_idx] + (self.block_size - self.idx) * value) / self.block_size
    self.idx = (self.idx + 1) % self.block_size
    if self.idx == 0:
      self.block_idx = (self.block_idx + 1) % self.num_blocks
      self.valid_blocks = min(self.valid_blocks + 1, self.num_blocks)

  def get(self):
    valid_block_idx = [i for i in range(self.valid_blocks) if i != self.block_idx]
    if not valid_block_idx:
      return None
    return np.mean(self.values[valid_block_idx], axis=0)


class LagEstimator:
  def __init__(self, CP, dt, block_count=10, block_size=100, window_sec=300.0, okay_window_sec=30.0,  min_vego=15, min_yr=np.radians(1), min_ncc=0.95):
    self.dt = dt
    self.window_sec = window_sec
    self.okay_window_sec = okay_window_sec
    self.initial_lag = CP.steerActuatorDelay + 0.2
    self.block_size = block_size
    self.block_count = block_count
    self.min_vego = min_vego
    self.min_yr = min_yr
    self.min_ncc = min_ncc

    self.t = 0
    self.lat_active = False
    self.steering_pressed = False
    self.desired_curvature = 0
    self.v_ego = 0
    self.yaw_rate = 0

    self.calibrator = PoseCalibrator()

    self.reset(self.initial_lag, 0)

  def reset(self, initial_lag, valid_blocks):
    window_len = int(self.window_sec / self.dt)
    self.points = Points(window_len)
    self.block_avg = BlockAverage(self.block_count, self.block_size, valid_blocks, initial_lag)
    self.lag = initial_lag

  def get_msg(self, valid):
    msg = messaging.new_message('liveActuatorDelay')

    msg.valid = valid

    liveDelay = msg.liveActuatorDelay
    liveDelay.steerActuatorDelay = self.lag
    liveDelay.isEstimated = self.block_avg.valid_blocks > 0
    liveDelay.validBlocks = self.block_avg.valid_blocks
    liveDelay.points = self.block_avg.values.tolist()

    return msg

  def handle_log(self, t, which, msg):
    if which == "carControl":
      self.lat_active = msg.latActive
    elif which == "carState":
      self.steering_pressed = msg.steeringPressed
      self.v_ego = msg.vEgo
    elif which == "controlsState":
      self.desired_curvature = msg.desiredCurvature
    elif which == "liveCalibration":
      self.calibrator.feed_live_calib(msg)
    elif which == "livePose":
      device_pose = Pose.from_live_pose(msg)
      calibrated_pose = self.calibrator.build_calibrated_pose(device_pose)
      self.yaw_rate = calibrated_pose.angular_velocity.z
    self.t = t

  def points_valid(self):
    return self.points.num_okay >= int(self.okay_window_sec / self.dt)

  def update_points(self):
    okay = self.lat_active and not self.steering_pressed and self.v_ego > self.min_vego and np.abs(self.yaw_rate) >= self.min_yr
    la_desired = self.desired_curvature * self.v_ego * self.v_ego
    la_actual_pose = self.yaw_rate * self.v_ego

    self.points.update(self.t, la_desired, la_actual_pose, okay)
    if not okay or not self.points_valid():
      return

    times, desired, actual, okay = self.points.get()
    times_interp = np.arange(times[-1] - self.window_sec, times[-1], DT_CTRL)
    desired_interp = np.interp(times_interp, times, desired)
    actual_interp = np.interp(times_interp, times, actual)
    okay_interp = np.interp(times_interp, times, okay).astype(bool)

    delay, corr = self.actuator_delay(desired_interp, actual_interp, okay_interp, DT_CTRL)
    if corr < self.min_ncc:
      return

    self.block_avg.update(delay)
    if (new_lag := self.block_avg.get()) is not None:
      self.lag = float(new_lag.item())

  def correlation_lags(self, sig_len, dt):
    return np.arange(0, sig_len) * dt

  def actuator_delay(self, expected_sig, actual_sig, mask, dt, max_lag=1.):
    """
    References:
      D. Padfield. "Masked FFT registration". In Proc. Computer Vision and
      Pattern Recognition, pp. 2918-2925 (2010).
      :DOI:`10.1109/CVPR.2010.5540032`
    """

    eps = np.finfo(np.float64).eps
    expected_sig = np.asarray(expected_sig, dtype=np.float64)
    actual_sig = np.asarray(actual_sig, dtype=np.float64)

    expected_sig[~mask] = 0.0
    actual_sig[~mask] = 0.0

    rotated_expected_sig = expected_sig[::-1]
    rotated_mask = mask[::-1]

    n = len(expected_sig) + len(actual_sig) - 1
    fft = partial(np.fft.fft, n=n)

    actual_sig_fft = fft(actual_sig)
    rotated_expected_sig_fft = fft(rotated_expected_sig)
    actual_mask_fft = fft(mask.astype(np.float64))
    rotated_mask_fft = fft(rotated_mask.astype(np.float64))

    number_overlap_masked_samples = np.fft.ifft(rotated_mask_fft * actual_mask_fft).real
    number_overlap_masked_samples[:] = np.round(number_overlap_masked_samples)
    number_overlap_masked_samples[:] = np.fmax(number_overlap_masked_samples, eps)
    masked_correlated_actual_fft = np.fft.ifft(rotated_mask_fft * actual_sig_fft).real
    masked_correlated_expected_fft = np.fft.ifft(actual_mask_fft * rotated_expected_sig_fft).real

    numerator = np.fft.ifft(rotated_expected_sig_fft * actual_sig_fft).real
    numerator -= masked_correlated_actual_fft * masked_correlated_expected_fft / number_overlap_masked_samples

    actual_squared_fft = fft(actual_sig ** 2)
    actual_sig_denom = np.fft.ifft(rotated_mask_fft * actual_squared_fft).real
    actual_sig_denom -= masked_correlated_actual_fft ** 2 / number_overlap_masked_samples
    actual_sig_denom[:] = np.fmax(actual_sig_denom, 0.0)

    rotated_expected_squared_fft = fft(rotated_expected_sig ** 2)
    expected_sig_denom = np.fft.ifft(actual_mask_fft * rotated_expected_squared_fft).real
    expected_sig_denom -= masked_correlated_expected_fft ** 2 / number_overlap_masked_samples
    expected_sig_denom[:] = np.fmax(expected_sig_denom, 0.0)

    denom = np.sqrt(actual_sig_denom * expected_sig_denom)

    # zero-out samples with very small denominators
    tol = 1e3 * eps * np.max(np.abs(denom), keepdims=True)
    nonzero_indices = denom > tol

    ncc = np.zeros_like(denom, dtype=np.float64)
    ncc[nonzero_indices] = numerator[nonzero_indices] / denom[nonzero_indices]
    np.clip(ncc, -1, 1, out=ncc)

    max_lag_samples = int(max_lag / dt)
    roi_ncc = ncc[len(expected_sig) - 1: len(expected_sig) - 1 + max_lag_samples]

    max_corr_index = np.argmax(roi_ncc)
    corr = roi_ncc[max_corr_index]
    lag = max_corr_index * dt

    return lag, corr


def main():
  config_realtime_process([0, 1, 2, 3], 5)

  pm = messaging.PubMaster(['liveActuatorDelay', 'alertDebug'])
  sm = messaging.SubMaster(['livePose', 'liveCalibration', 'carControl', 'carState', 'controlsState'], poll='livePose')

  params = Params()
  CP = messaging.log_from_bytes(params.get("CarParams", block=True), car.CarParams)
  estimator = LagEstimator(CP, 1. / SERVICE_LIST['livePose'].frequency)

  lag_params = params.get("LiveLag")
  if lag_params:
    try:
      with log.Event.from_bytes(lag_params) as msg:
        lag_init = msg.liveActuatorDelay.steerActuatorDelay
        valid_blocks = msg.liveActuatorDelay.validBlocks
        estimator.reset(lag_init, valid_blocks)
    except Exception:
      print("Error reading cached LagParams")

  while True:
    sm.update()
    if sm.all_checks():
      for which in sorted(sm.updated.keys(), key=lambda x: sm.logMonoTime[x]):
        if sm.updated[which]:
          t = sm.logMonoTime[which] * 1e-9
          estimator.handle_log(t, which, sm[which])
      estimator.update_points()

    if sm.frame % 25 == 0:
      msg = estimator.get_msg(sm.all_checks())
      alert_msg = messaging.new_message('alertDebug')
      alert_msg.alertDebug.alertText1 = f"Lag estimate (fixed: {CP.steerActuatorDelay:.2f} s)"
      alert_msg.alertDebug.alertText2 = f"{msg.liveActuatorDelay.steerActuatorDelay:.2f} s ({msg.liveActuatorDelay.isEstimated})"

      pm.send('liveActuatorDelay', msg)
      pm.send('alertDebug', alert_msg)

      if msg.liveActuatorDelay.isEstimated: # TODO maybe to often once estimated
        params.put_nonblocking("LiveLag", msg.to_bytes())


if __name__ == "__main__":
  main()