SigPack - the C++ signal processing library
sp::UKF Class Reference

#include <kalman.h>

Inheritance diagram for sp::UKF:
Inheritance graph
Collaboration diagram for sp::UKF:
Collaboration graph

Public Member Functions

 UKF (arma::uword _N, arma::uword _M, arma::uword _L)
 
void set_alpha (double _a)
 
void set_beta (double _b)
 
void set_kappa (double _k)
 
void set_lambda (double _l)
 
void update_weights (void)
 
void update_sigma (const arma::mat &_x, const arma::mat &_P)
 
arma::mat ut (const arma::mat &_x, const arma::mat &_P, const fcn_v _f)
 
void predict (const arma::mat u)
 
void predict (void)
 
void update (const arma::mat z)
 
void rts_smooth (const arma::mat &Xf, const arma::cube &Pf, arma::mat &Xs, arma::cube &Ps)
 
- Public Member Functions inherited from sp::KF
 KF (arma::uword _N, arma::uword _M, arma::uword _L)
 
 ~KF ()
 
void clear (void)
 
void set_state_vec (const arma::mat &_x)
 
void set_trans_mat (const arma::mat &_A)
 
void set_control_mat (const arma::mat &_B)
 
void set_meas_mat (const arma::mat &_H)
 
void set_err_cov (const arma::mat &_P)
 
void set_proc_noise (const arma::mat &_Q)
 
void set_meas_noise (const arma::mat &_R)
 
void set_kalman_gain (const arma::mat &_K)
 
void set_trans_fcn (fcn_v _f)
 
void set_meas_fcn (fcn_v _h)
 
arma::mat get_state_vec (void)
 
arma::mat get_err (void)
 
arma::mat get_kalman_gain (void)
 
arma::mat get_err_cov (void)
 
void predict (const arma::mat u)
 
void predict (void)
 
void update (const arma::mat z)
 
void rts_smooth (const arma::mat &Xf, const arma::cube &Pf, arma::mat &Xs, arma::cube &Ps)
 

Protected Attributes

double alpha
 
double beta
 
double kappa
 
double lambda
 
arma::mat X
 
arma::mat S
 
arma::mat C
 
arma::vec Wx
 
arma::vec Wp
 
- Protected Attributes inherited from sp::KF
arma::uword N
 
arma::uword M
 
arma::uword L
 
bool lin_proc
 
bool lin_meas
 
arma::mat x
 
arma::mat z_err
 
arma::mat A
 
arma::mat B
 
arma::mat H
 
arma::mat P
 
arma::mat Q
 
arma::mat R
 
arma::mat K
 
fcn_v f
 
fcn_v h
 

Detailed Description

Implements Kalman functions for the discrete system with additive noise

\[ x_k = f(x_{k-1})+Bu_{k-1} + w_{k-1} \]

and with measurements

\[ z_k = h(x_k) + v_k \]

where f(x) and h(x) may be nonlinear functions.

The predict and update stage is using the unscented transform of the sigma points of the input and/or the measurements.
For detailed info see: http://www.lce.hut.fi/~ssarkka/course_k2011/pdf/course_booklet_2011.pdf

Examples:
kalman_UKF.cpp.

Definition at line 563 of file kalman.h.


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