auterion::LosFilter Class Reference

A filter for estimating Line-of-Sight (LOS) vectors and their rates. More...

#include <auterion_sdk/tracking/los_filter/los_filter.hpp>

Public Member Functions
	LosFilter ()
	Default constructor.
void	init ()
	Initializes the filter without an initial LOS vector.
void	init (const Eigen::Vector3f &los_init, float los_std_init=0.1f, float los_rate_std_init=0.1f)
	Initializes the filter with a given LOS vector.
bool	isInitialized () const
	Checks whether the filter has been initialized.
void	reset ()
	Resets the filter to the uninitialized state, clearing state and covariance.
void	predict (float dt, float los_rate_process_nsd=0.1f)
	Predicts the next state of the filter based on elapsed time.
void	update (const Eigen::Vector3f &los_measurement, float los_variance)
	Updates the filter state with a new LOS measurement.
void	update (const Eigen::Vector3f &los_measurement, const Eigen::Matrix3f &measurement_cov)
	Updates the filter state with a new LOS measurement and a full measurement covariance.
void	resetLos (const Eigen::Vector3f &los, float los_std=0.1f)
	Moves the LOS estimate while keeping the estimated LOS rate.
Eigen::Vector3f	getLOS () const
	Gets the current estimated LOS vector.
Eigen::Vector3f	getLOSRate () const
	Gets the current estimated LOS rate.
const Eigen::MatrixXf	getCovariance () const
	Gets the covariance matrix of the state.
Eigen::Matrix< float, 3, 2 >	getTangentBasis () const
	Gets the world-frame tangent-plane basis of the LOS.
Eigen::Matrix2f	getThetaCovariance () const
	Gets the covariance of the LOS error states, in the getTangentBasis() frame.
Eigen::Vector3f	getLOSInnov () const
	Gets the LOS innovation (difference between measurement and prediction).
Eigen::Vector3f	getLOSInnovVar () const
	Gets the variance of the LOS innovation.
Eigen::MatrixXf	getStateVariance () const
	Gets the state variance matrix.
Eigen::MatrixXf	getLOSVariance () const
	Gets the variance of the LOS estimates.
Eigen::MatrixXf	getLOSRateVariance () const
	Gets the variance of the LOS rate estimates.

Detailed Description

A filter for estimating Line-of-Sight (LOS) vectors and their rates.

This class provides functionality for initializing, predicting, and updating the state of a LOS estimation filter. It maintains covariance information and provides access to various state variables.

Constructor & Destructor Documentation

LosFilter()

auterion::LosFilter::LosFilter

(

)

Default constructor.

Initializes an instance of the LosFilter.

Member Function Documentation

getCovariance()

const Eigen::MatrixXf auterion::LosFilter::getCovariance

(

)

const

Gets the covariance matrix of the state.

Returns

State covariance matrix.

getLOS()

Eigen::Vector3f auterion::LosFilter::getLOS

(

)

const

Gets the current estimated LOS vector.

Returns

Estimated LOS vector.

getLOSInnov()

Eigen::Vector3f auterion::LosFilter::getLOSInnov

(

)

const

Gets the LOS innovation (difference between measurement and prediction).

Returns

LOS innovation vector.

getLOSInnovVar()

Eigen::Vector3f auterion::LosFilter::getLOSInnovVar

(

)

const

Gets the variance of the LOS innovation.

Returns

LOS innovation variance vector.

getLOSRate()

Eigen::Vector3f auterion::LosFilter::getLOSRate

(

)

const

Gets the current estimated LOS rate.

Returns

Estimated LOS rate vector.

getLOSRateVariance()

Eigen::MatrixXf auterion::LosFilter::getLOSRateVariance

(

)

const

Gets the variance of the LOS rate estimates.

Returns

LOS rate variance matrix.

getLOSVariance()

Eigen::MatrixXf auterion::LosFilter::getLOSVariance

(

)

const

Gets the variance of the LOS estimates.

Returns

LOS variance matrix.

getStateVariance()

Eigen::MatrixXf auterion::LosFilter::getStateVariance

(

)

const

Gets the state variance matrix.

Returns

State variance matrix.

getTangentBasis()

Eigen::Matrix< float, 3, 2 > auterion::LosFilter::getTangentBasis

(

)

const

Gets the world-frame tangent-plane basis of the LOS.

The two columns are the unit directions the LOS error states perturb the LOS along: both orthogonal to getLOS() and right-handed with it. Together with getThetaCovariance() they give the LOS direction covariance T * C * T^T. This is the filter's own tangent frame and cannot be reconstructed from getLOS() alone.

Returns

3x2 matrix whose columns span the plane orthogonal to the LOS.

getThetaCovariance()

Eigen::Matrix2f auterion::LosFilter::getThetaCovariance

(

)

const

Gets the covariance of the LOS error states, in the getTangentBasis() frame.

Returns

2x2 LOS error covariance (rad^2).

init()

void auterion::LosFilter::init	(	const Eigen::Vector3f &	los_init,
		float	los_std_init = 0.1f,
		float	los_rate_std_init = 0.1f
	)

Initializes the filter with a given LOS vector.

Parameters

los_init	Initial Line-of-Sight (LOS) vector.
los_std_init	Initial LOS state standard deviation (rad). Defaults to 0.1.
los_rate_std_init	Initial LOS rate state standard deviation (rad/s). Defaults to 0.1.

isInitialized()

bool auterion::LosFilter::isInitialized

(

)

const

Checks whether the filter has been initialized.

Returns

True if init() has been called and reset() has not been called since.

predict()

void auterion::LosFilter::predict	(	float	dt,
		float	los_rate_process_nsd = 0.1f
	)

Predicts the next state of the filter based on elapsed time.

Parameters

dt	Time step for prediction (in seconds).
los_rate_process_nsd	LOS rate noise spectral density in rad/(s * sqrt(Hz)). Defaults to 0.1.

resetLos()

void auterion::LosFilter::resetLos	(	const Eigen::Vector3f &	los,
		float	los_std = 0.1f
	)

Moves the LOS estimate while keeping the estimated LOS rate.

Applies the minimal world-frame rotation onto the given LOS, which carries the rate estimate with it, re-seeds the LOS variance and clears the LOS/rate cross-covariance. The rate variance is left untouched. Intended for small anchor corrections, where a full init() would throw away a rate estimate that is still valid. Behaves as init() if the filter has not been initialized.

Parameters

los	New LOS vector.
los_std	LOS state standard deviation to re-seed with (rad). Defaults to 0.1.

update() [1/2]

void auterion::LosFilter::update	(	const Eigen::Vector3f &	los_measurement,
		const Eigen::Matrix3f &	measurement_cov
	)

Updates the filter state with a new LOS measurement and a full measurement covariance.

Anisotropic measurement noise is weighted along its actual direction instead of inflating every axis. Use this when the dominant error has a known direction, for example an attitude-rate error (omega x los) * dt.

Parameters

los_measurement	Measured LOS vector.
measurement_cov	Measurement covariance in world axes. Must be positive definite: the component along the LOS itself is unobservable and only keeps the update well conditioned. A covariance that is not positive definite leaves the filter untouched.

update() [2/2]

void auterion::LosFilter::update	(	const Eigen::Vector3f &	los_measurement,
		float	los_variance
	)

Updates the filter state with a new LOS measurement.

Parameters

los_measurement	Measured LOS vector.
los_variance	Measurement variance.

---

The documentation for this class was generated from the following file:

• include/auterion_sdk/tracking/los_filter/los_filter.hpp