jafar::filter::ParticleFilter Class Reference

Basic particle filter. More...

---

Detailed Description

Basic particle filter.

Definition at line 18 of file particleFilter.hpp.

#include <particleFilter.hpp>

List of all members.

Public Types
typedef std::vector < jblas::vec >::iterator	STATE_IT
typedef std::vector < jblas::vec >::const_iterator	STATE_CONST_IT
Public Member Functions
	ParticleFilter (std::size_t nbParticles_, std::size_t sizeState_, const jblas::vec &stateMin_, const jblas::vec &stateMax_)
	Create a particle filter.
std::size_t	sizeState () const
std::size_t	getNbParticles () const
double	getWeight (std::size_t index_) const
const jblas::vec &	getState (std::size_t index_) const
void	predict (LinearPredictModel &m_)
void	update (jafar::filter::JacobianObserveModel &m_, double z_)
	This function computes the observation step (innovation is given).
const jblas::vec &	getStateMean ()
	Resample the particles.
const jblas::sym_mat &	getStateCov ()
Protected Member Functions
void	computeMeanAndCov ()
Protected Attributes
std::size_t	nbParticles
std::size_t	_sizeState
jblas::vec	weight
std::vector< jblas::vec >	states
jblas::vec	stateMean
jblas::sym_mat	stateCov
bool	meanAndCovUpToDate
jafar::jmath::MultiDimNormalDistribution	predictNoise

---

Constructor & Destructor Documentation

jafar::filter::ParticleFilter::ParticleFilter	(	std::size_t	nbParticles_,
		std::size_t	sizeState_,
		const jblas::vec &	stateMin_,
		const jblas::vec &	stateMax_
	)

Create a particle filter.

Parameters:

nbParticles_	the initial number of particles
sizeState_	the size of the estimated state Create a particle filter with a set of identical particles.
nbParticles_	The initial number of particles
initState_	The initial state of the particles Create a particule filter with uniformly chosen particles.
nbParticles_	The initial number of particles
sizeState_	the size of the estimated state
stateMin_	The lower state
stateMax_	The upper state

---

Member Function Documentation

const jblas::sym_mat& jafar::filter::ParticleFilter::getStateCov

(

)

Returns:

The covariance matrix of the state estimated by the particles

const jblas::vec& jafar::filter::ParticleFilter::getStateMean

(

)

Resample the particles.

The number of particles remains the same. Resample a new nember of particles.

Parameters:

nbParticles_

The new nember of particles

void jafar::filter::ParticleFilter::update	(	jafar::filter::JacobianObserveModel &	m_,
		double	z_
	)

This function computes the observation step (innovation is given).

Parameters:

m_	observation model
z_	observation
R_	measure noise covariance

---

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

• modules/filter/include/filter/particleFilter.hpp