ADCS.estimators.attitude_estimators.attitude_AugmentedEKF module

Additive EKF with joint bias and disturbance-parameter estimation.

1. Construct the augmented full-state estimate

Validate the physical state and append the configured actuator-bias, sensor-bias, and disturbance-parameter blocks to the full-quaternion covariance layout.

2. Propagate the augmented nominal state

Synchronize the nominal augmented parameters into EstimatedSatellite and propagate the spacecraft state through the nonlinear dynamics.

3. Linearize the augmented process model

Build the physical, bias, and disturbance-parameter couplings and discretize the continuous process noise.

4. Predict the full covariance

Apply \(P^- = \\Phi P^+ \\Phi^T + Q_d\) in full quaternion coordinates.

5. Predict measurements and form the innovation

Use the shared measurement stack, including bias terms in the augmented measurement model.

6. Form the augmented measurement Jacobian

Include derivatives with respect to the physical state, sensor biases, and disturbance parameters.

7. Apply the Kalman correction

Update the complete augmented estimate and covariance.

8. Normalize and retain the augmented state

Normalize the quaternion and retain the estimated parameter blocks for the next prediction.

class ADCS.estimators.attitude_estimators.attitude_AugmentedEKF.AugmentedEKF(satellite, state, *, dt, unmodeled_dynamics_psd=0.0, measurement_quaternion_mode='quaternion_vector')[source]

Bases: EKF

Additive EKF for the full EstimatorState layout.

With empty augmented blocks this is equivalent to the existing EKF. When the estimated satellite and state provide bias or disturbance parameter blocks, the shared full-quaternion covariance includes them in the prediction and measurement update.

Parameters:

  • satellite (Any)

  • state (EstimatorState)

  • dt (float)

  • unmodeled_dynamics_psd (Any)

  • measurement_quaternion_mode (str)

supports_augmented_parameters = True