def run_tracker(): # parameters filter_misestimation_factor = 1.0 sample_size = 100 used_taps = int(sample_size * 0.5) measurement_std = 3.5 # set up sensor simulator dt = 0.1 measurement_std_list = np.asarray([measurement_std] * sample_size) sim = SensorSim(0, 0.1, measurement_std_list, 1, timestep=dt) # set up kalman filter tracker = KalmanFilter(dim_x=2, dim_z=1) tracker.F = np.array([[1, dt], [0, 1]]) q = Q_discrete_white_noise(dim=2, dt=dt, var=0.01) tracker.Q = q tracker.H = np.array([[1, 0]]) tracker.R = measurement_std**2 * filter_misestimation_factor tracker.x = np.array([[0, 0]]).T tracker.P = np.eye(2) * 500 # perform sensor simulation and filtering readings = [] truths = [] mu = [] residuals = [] for _ in measurement_std_list: reading, truth = sim.read() readings.extend(reading.flatten()) truths.extend(truth.flatten()) tracker.predict() tracker.update(reading) mu.extend(tracker.x[0]) residual_posterior = reading - np.dot(tracker.H, tracker.x) residuals.extend(residual_posterior[0]) # error = np.asarray(truths) - mu # plot_results(readings, mu, error, residuals) # perform estimation cor = Correlator(residuals) correlation = cor.autocorrelation(used_taps) R_approx = estimate_noise_approx(correlation[0], tracker.H, tracker.P, 'posterior') abs_err = measurement_std**2 - R_approx rel_err = abs_err / measurement_std**2 print("True: %.3f" % measurement_std**2) print("Filter: %.3f" % tracker.R) print("Estimated (approximation): %.3f" % R_approx) print("Absolute error: %.3f" % abs_err) print("Relative error: %.3f %%" % (rel_err * 100)) print("-" * 15) return rel_err
class TestSensorSim: def setup(self): self.measurement_std = [1, 2, 3] self.sim1d = SensorSim(0, 0.5, self.measurement_std, 1) self.sim2d = SensorSim((0, 0), (1, 2), self.measurement_std, 2) def test_basic_constructor(self): pass def test_constructor_measurement_check(self): with pytest.raises(ValueError): SensorSim(0, 1, 4, 1) def test_constructor_position_check(self): with pytest.raises(ValueError): SensorSim(0, (1, 1), [4], 2) def test_constructor_velocity_check(self): with pytest.raises(ValueError): SensorSim((0, 0), 1, [4], 2) def test_read_all_1d(self): for x in self.measurement_std: meas, truth = self.sim1d.read() assert meas.shape == (1, 1) assert truth.shape == (1, 1) def test_read_all_2d(self): for x in self.measurement_std: meas, truth = self.sim2d.read() assert meas.shape == (2, 1) assert truth.shape == (2, 1) def test_read_too_many(self): for x in self.measurement_std: self.sim2d.read() with pytest.raises(IndexError): self.sim2d.read() def test_batch_read_2d(self): res = self.sim2d.batch_read() assert len(res) == len(self.measurement_std) shape = np.asarray(res).shape assert shape[0] == len(self.measurement_std) assert shape[1] == 2 assert shape[2] == 2 assert shape[3] == 1
""" Evaluate the SensorSim class by plotting the result of an increasing and later decreasing measurement noise variance """ import numpy as np from matplotlib import pyplot as plt from noiseestimation.sensorsim import SensorSim measurement_std = np.arange(0, 4, .02) measurement_std = np.concatenate( (measurement_std, list(reversed(measurement_std)))) sim = SensorSim( position=( 0, 0), velocity=( 0.5, 1), measurement_std=measurement_std, dim=2) readings = np.array([np.array(sim.read()) for _ in measurement_std]) plt.plot(readings[:, 0, 0], readings[:, 0, 1], 'go', label="sensor readings") plt.plot(readings[:, 1, 0], readings[:, 1, 1], 'm', linewidth=2, label="true values") plt.legend(loc='lower right') plt.title("Position sensor readings") plt.show()
def run_tracker(): # parameters measurement_std = 3.5 filter_misestimation_factor = 1.0 sample_size = 2000 estimation_sample_size = 80 used_taps = int(estimation_sample_size * 0.5) # set up sensor simulator dt = 0.1 # measurement_std_list = np.asarray([measurement_std] * sample_size) measurement_std_list = np.linspace(measurement_std / 5, measurement_std * 1, sample_size / 2) measurement_std_list = np.concatenate( (measurement_std_list, list(reversed(measurement_std_list)))) sim = SensorSim(0, 0.1, measurement_std_list, 1, timestep=dt) # set up kalman filter tracker = KalmanFilter(dim_x=2, dim_z=1) tracker.F = np.array([[1, dt], [0, 1]]) q = Q_discrete_white_noise(dim=2, dt=dt, var=0.01) tracker.Q = q tracker.H = np.array([[1, 0]]) tracker.R = measurement_std**2 * filter_misestimation_factor tracker.x = np.array([[0, 0]]).T tracker.P = np.eye(2) * 500 # perform sensor simulation and filtering readings = [] truths = [] mu = [] residuals = [] Rs = [] for idx, _ in enumerate(measurement_std_list): reading, truth = sim.read() readings.extend(reading.flatten()) truths.extend(truth.flatten()) tracker.predict() tracker.update(reading) mu.extend(tracker.x[0]) residuals.extend(tracker.y[0]) Rs.append(tracker.R) if (idx < estimation_sample_size or idx % (estimation_sample_size / 10) != 0): print(idx) continue cor = Correlator(residuals[-estimation_sample_size:]) used_taps = int(estimation_sample_size / 2) correlation = cor.autocorrelation(used_taps) R = estimate_noise(correlation, tracker.K, tracker.F, tracker.H) R_approx = estimate_noise_approx(correlation[0], tracker.H, tracker.P) abs_err = measurement_std**2 - R rel_err = abs_err / measurement_std**2 print("True: %.3f" % measurement_std**2) print("Filter: %.3f" % tracker.R) print("Estimated: %.3f" % R) print("Estimated (approximation): %.3f" % R_approx) print("Absolute error: %.3f" % abs_err) print("Relative error: %.3f %%" % (rel_err * 100)) print("-" * 15) if (R > 0): tracker.R = R # if(R_approx > 0): # tracker.R = R_approx # error = np.asarray(truths) - mu plot_results(readings, mu, Rs, measurement_std_list)