#include <KalmanFilter.hpp>

Public Member Functions
EIGEN_MAKE_ALIGNED_OPERATOR_NEW	KF ()

	~KF ()

void	predictionDiscrete (Eigen::Matrix< double, SIZE, SIZE > F, Eigen::Matrix< double, SIZE, SIZE > Q, double dt)

template<unsigned int INPUT_SIZE>
void	prediction (Eigen::Matrix< double, INPUT_SIZE, 1 > u, Eigen::Matrix< double, SIZE, INPUT_SIZE > B, Eigen::Matrix< double, SIZE, SIZE > F, Eigen::Matrix< double, SIZE, SIZE > Q)

template<unsigned int MEASUREMENT_SIZE>
void	correction (Eigen::Matrix< double, MEASUREMENT_SIZE, 1 > p, Eigen::Matrix< double, MEASUREMENT_SIZE, SIZE > H, Eigen::Matrix< double, MEASUREMENT_SIZE, MEASUREMENT_SIZE > R)

template<unsigned int MEASUREMENT_SIZE>
Eigen::Matrix< double, MEASUREMENT_SIZE, 1 >	innovation (Eigen::Matrix< double, MEASUREMENT_SIZE, 1 > p, Eigen::Matrix< double, MEASUREMENT_SIZE, SIZE > H)

template<unsigned int MEASUREMENT_SIZE>
Eigen::Matrix< double, MEASUREMENT_SIZE, MEASUREMENT_SIZE >	innovationCovariance (Eigen::Matrix< double, MEASUREMENT_SIZE, SIZE > H, Eigen::Matrix< double, MEASUREMENT_SIZE, MEASUREMENT_SIZE > R)

template<unsigned int MEASUREMENT_SIZE>
void	update (Eigen::Matrix< double, MEASUREMENT_SIZE, SIZE > H, Eigen::Matrix< double, SIZE, MEASUREMENT_SIZE > K, Eigen::Matrix< double, MEASUREMENT_SIZE, 1 > y)

template<unsigned int MEASUREMENT_SIZE>
Eigen::Matrix< double, SIZE, MEASUREMENT_SIZE >	gain (Eigen::Matrix< double, MEASUREMENT_SIZE, SIZE > H, Eigen::Matrix< double, MEASUREMENT_SIZE, MEASUREMENT_SIZE > S)

template<unsigned int MEASUREMENT_SIZE, unsigned int DEGREE_OF_FREEDOM>
bool	correctionChiSquare (const Eigen::Matrix< double, MEASUREMENT_SIZE, 1 > &p, const Eigen::Matrix< double, MEASUREMENT_SIZE, MEASUREMENT_SIZE > &R, const Eigen::Matrix< double, MEASUREMENT_SIZE, SIZE > &H, float chi_square_reject_threshold)

Public Attributes
Eigen::Matrix< double, SIZE, 1 >	x

Eigen::Matrix< double, SIZE, SIZE >	P

pose_ekf::ChiSquared *	chi_square

Constructor & Destructor Documentation

template<unsigned int SIZE>

EIGEN_MAKE_ALIGNED_OPERATOR_NEW KalmanFilter::KF< SIZE >::KF ( )

inline

what is this line

template<unsigned int SIZE>

KalmanFilter::KF< SIZE >::~KF ( )

inline

Member Function Documentation

template<unsigned int SIZE>

template<unsigned int MEASUREMENT_SIZE>

void KalmanFilter::KF< SIZE >::correction	(	Eigen::Matrix< double, MEASUREMENT_SIZE, 1 >	p,
		Eigen::Matrix< double, MEASUREMENT_SIZE, SIZE >	H,
		Eigen::Matrix< double, MEASUREMENT_SIZE, MEASUREMENT_SIZE >	R
	)

inline

correction step, p - observation H - observation function R - measurement noise covariance matrix

template<unsigned int SIZE>

template<unsigned int MEASUREMENT_SIZE, unsigned int DEGREE_OF_FREEDOM>

bool KalmanFilter::KF< SIZE >::correctionChiSquare	(	const Eigen::Matrix< double, MEASUREMENT_SIZE, 1 > &	p,
		const Eigen::Matrix< double, MEASUREMENT_SIZE, MEASUREMENT_SIZE > &	R,
		const Eigen::Matrix< double, MEASUREMENT_SIZE, SIZE > &	H,
		float	chi_square_reject_threshold
	)

inline

Applies the Kalman Filter correction using a chi_square method to reject bad observation samples p - Observation R - observation noise H - how state is translated into observation chi_square_reject_threshold - threashold for the rejection (0 = no rejection) return bool - if the observation was rejected or not.

template<unsigned int SIZE>

template<unsigned int MEASUREMENT_SIZE>

Eigen::Matrix<double, SIZE, MEASUREMENT_SIZE> KalmanFilter::KF< SIZE >::gain	(	Eigen::Matrix< double, MEASUREMENT_SIZE, SIZE >	H,
		Eigen::Matrix< double, MEASUREMENT_SIZE, MEASUREMENT_SIZE >	S
	)

inline

calculates the gain J_H - jacobian observation model

template<unsigned int SIZE>

template<unsigned int MEASUREMENT_SIZE>

Eigen::Matrix<double, MEASUREMENT_SIZE, 1> KalmanFilter::KF< SIZE >::innovation	(	Eigen::Matrix< double, MEASUREMENT_SIZE, 1 >	p,
		Eigen::Matrix< double, MEASUREMENT_SIZE, SIZE >	H
	)

inline

template<unsigned int SIZE>

template<unsigned int MEASUREMENT_SIZE>

Eigen::Matrix<double, MEASUREMENT_SIZE, MEASUREMENT_SIZE> KalmanFilter::KF< SIZE >::innovationCovariance	(	Eigen::Matrix< double, MEASUREMENT_SIZE, SIZE >	H,
		Eigen::Matrix< double, MEASUREMENT_SIZE, MEASUREMENT_SIZE >	R
	)

inline

innovation covariance step, p - observation h - observation model function H - observation model R - measurement noise covariance matrix

template<unsigned int SIZE>

template<unsigned int INPUT_SIZE>

void KalmanFilter::KF< SIZE >::prediction	(	Eigen::Matrix< double, INPUT_SIZE, 1 >	u,
		Eigen::Matrix< double, SIZE, INPUT_SIZE >	B,
		Eigen::Matrix< double, SIZE, SIZE >	F,
		Eigen::Matrix< double, SIZE, SIZE >	Q
	)

inline

update step u - Input B - relation between input and state F - state transition Q - process noise covariance matrix,

template<unsigned int SIZE>

void KalmanFilter::KF< SIZE >::predictionDiscrete	(	Eigen::Matrix< double, SIZE, SIZE >	F,
		Eigen::Matrix< double, SIZE, SIZE >	Q,
		double	dt
	)

inline

prediction step F - state transition Q - process noise covariance matrix, dt - discrete time time step

template<unsigned int SIZE>

template<unsigned int MEASUREMENT_SIZE>

void KalmanFilter::KF< SIZE >::update	(	Eigen::Matrix< double, MEASUREMENT_SIZE, SIZE >	H,
		Eigen::Matrix< double, SIZE, MEASUREMENT_SIZE >	K,
		Eigen::Matrix< double, MEASUREMENT_SIZE, 1 >	y
	)

inline

update step, H - observation model y - innovation K - gain

Member Data Documentation

template<unsigned int SIZE>

pose_ekf::ChiSquared* KalmanFilter::KF< SIZE >::chi_square

fault detection libary

template<unsigned int SIZE>

Eigen::Matrix<double, SIZE, SIZE> KalmanFilter::KF< SIZE >::P

covariance matrix

template<unsigned int SIZE>

Eigen::Matrix<double,SIZE,1> KalmanFilter::KF< SIZE >::x

do i need to declare the size as dynamic ? state estimate

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

/build/rock-master-18.09-slam-pose-ekf-0.20121031/src/KalmanFilter.hpp

Public Member Functions

Public Attributes

Constructor & Destructor Documentation

Member Function Documentation

Member Data Documentation