def setup():
sim = PlaybackSensor("data/vehicle_state.json",
["fStwAng", "fVx", "fYawrate"])
# set up kalman filter
tracker = BicycleEKF_noVel(dt)
tracker.R = sim_var + measurement_var
tracker.x = np.array([[0, 0]]).T
tracker.P = np.eye(2) * 500
return sim, tracker
def setup():
sim = PlaybackSensor("data/vehicle_state_integrated.json",
["fVx", "fStwAng", "pose_x", "pose_y"])
# set up kalman filter
tracker = SimpleBicycleEKF(dt, wheelbase, var_vel, var_steer)
tracker.Q = np.eye(2) * 0.001
tracker.H = np.array([[1, 0, 0], [0, 1, 0]])
tracker.R = var_measurement
tracker.x = np.array([[0, 0, 0]]).T
tracker.P = np.eye(3) * 500
return sim, tracker
def setup():
sim = PlaybackSensor("data/vehicle_state_adenauer.json",
fields=["fYawrate", "fVx"],
control_fields=["fStwAng", "fAx"])
# set up kalman filter
tracker = BicycleUKF(dt)
Q_factor = np.array(1e-5)
tracker.Q = np.diag([1, 1, 1] * Q_factor)
tracker.R = R_proto * (sim_var + measurement_var) * 1
tracker.x = np.array([0, 0, 1]).T
tracker.P = np.eye(3) * 1e0
return sim, tracker
def setup():
sim = PlaybackSensor("data/vehicle_state.json",
fields=["fYawrate", "fVx"],
control_fields=["fStwAng", "fAx"])
# set up kalman filter
tracker = BicycleUKF(dt)
Q_factor = np.array(5e-7)
tracker.Q = np.diag([1, 1, 0.5] * Q_factor)
tracker.R = R_proto * (sim_var) * filter_misestimation_factor
tracker.x = np.array([0, 0, 1]).T
tracker.P = np.eye(3) * 1e0
return sim, tracker
def setup():
sim = PlaybackSensor("data/vehicle_state.json",
fields=["fYawrate", "fVx"],
control_fields=["fStwAng", "fAx"])
# set up kalman filter
tracker = BicycleEKF(dt)
tracker.R = sim_var + measurement_var
tracker.x = np.array([[0, 0, 1e-3]]).T
tracker.P = np.eye(3) * 500
tracker.var_steer = var_steer
tracker.var_acc = var_acc
return sim, tracker
def setup():
sim = PlaybackSensor("data/vehicle_state.json",
fields=["fYawrate", "fVx"],
control_fields=["fStwAng", "fAx"])
# set up kalman filter
tracker = BicycleEKF(dt)
tracker.R = R_proto * (sim_var + measurement_var)
tracker.x = np.array([[0, 0, 1e-3]]).T
tracker.P = np.eye(3) * 500
# control not as accurate anymore
tracker.var_steer = 0.0005
tracker.var_acc = 0.5
return sim, tracker
def setup():
sim = PlaybackSensor("data/vehicle_state.json",
fields=["fYawrate", "fVx"],
control_fields=["fStwAng", "fAx"])
# set up kalman filter
tracker = BicycleUKFBias(dt)
Q_factor = np.array(5e-7)
tracker.Q = np.diag([1, 1, 1, 1e-11] * Q_factor)
tracker.R = R_proto * (sim_var + measurement_var) * filter_misestimation_factor
tracker.x = np.array([0, 0, 1, 0.35 * np.pi / 180]).T
tracker.P = np.eye(4) * 1e0
tracker.P[3, 3] = 1e-8
return sim, tracker
def setup():
F = generate_F_matrix(velocity=0.001)
H = np.array([[0, 1]])
sim = PlaybackSensor("data/vehicle_state.json",
["fVx", "fYawrate", "fStwAng"])
# set up kalman filter
tracker = KalmanFilter(dim_x=2, dim_z=1)
tracker.F = F
tracker.Q = np.eye(2) * 0.001
tracker.H = H
tracker.R = measurement_var
tracker.x = np.array([[0, 0]]).T
tracker.P = np.eye(2) * 500
return sim, tracker
def setup():
sim = PlaybackSensor("data/vehicle_state.json",
fields=["fYawrate", "fVx"],
control_fields=["fStwAng", "fAx"])
# set up kalman filter
tracker = BicycleEKF(dt)
tracker.R = R_proto * (sim_var + measurement_var) * \
filter_misestimation_factor
tracker.x = np.array([[0, 0, 1e-3]]).T
tracker.P = np.eye(3) * 500
# control not as accurate anymore
tracker.var_steer = 0
tracker.var_acc = 0
Q_factor = np.array(5e-7)
tracker.Q = np.diag([1, 1, 0.5] * Q_factor)
return sim, tracker