ADCS.estimators.attitude_estimators.attitude_AugmentedMEKF module

Multiplicative EKF with joint bias and disturbance-parameter estimation.

1. Construct the augmented tangent-state estimate

Validate the physical state and append bias and disturbance-parameter blocks while retaining the three-coordinate right-attitude error.

2. Propagate the augmented nominal state

Synchronize the nominal augmented parameters into EstimatedSatellite and propagate the nonlinear spacecraft model.

3. Linearize the augmented process model

Construct the tangent error dynamics and discretize physical, bias, and disturbance-parameter process noise.

4. Predict the tangent covariance

Apply the linear covariance recursion in the right-error coordinates.

5. Predict measurements and form the innovation

Use the shared measurement stack and its bias-aware residual definitions.

6. Form the augmented measurement Jacobian

Include physical-state, sensor-bias, and disturbance-parameter sensitivities.

7. Apply the Kalman correction

Correct the physical state and every enabled augmented parameter block.

8. Retract the attitude and reset the tangent covariance

Apply the multiplicative quaternion retraction and transport the covariance about the new tangent origin.

class ADCS.estimators.attitude_estimators.attitude_AugmentedMEKF.AugmentedMEKF(satellite, state, *, dt, unmodeled_dynamics_psd=0.0, quaternion_mode='quaternion_vector')[source]

Bases: MEKF

MEKF for the full EstimatorState layout.

The state may contain actuator biases, attitude-sensor biases, and estimated disturbance parameters in addition to angular velocity, attitude, and reaction-wheel momentum. Their dynamics and measurement couplings are supplied by EstimatedSatellite and the shared process model/measurement stack; the covariance remains in right tangent coordinates, just as for MEKF.

Parameters:

  • satellite (Any)

  • state (EstimatorState)

  • dt (float)

  • unmodeled_dynamics_psd (Any)

  • quaternion_mode (str)

supports_augmented_parameters = True