RCAC Class Reference

#include <RCAC.hpp>

Inheritance diagram for RCAC:

Collaboration diagram for RCAC:

Public Member Functions
void	oneStep (Eigen::VectorXd &uIn, Eigen::VectorXd &zIn, Eigen::VectorXd &yIn)
Eigen::VectorXd	getControl ()
Eigen::VectorXd	getCoeff ()
int	getlu ()
int	getly ()
int	getlz ()
int	getNc ()
int	getkk ()

Static Public Member Functions
template<typename T >
static RCAC *	init (T &FLAGS, rcacFilt &FILT, std::string &whichRCAC)

Protected Member Functions
virtual void	coeffUpdate (Eigen::VectorXd &zIn)=0
void	initRegressor ()
void	initFiltered ()
void	computeFiltered ()

Protected Attributes
int	lz
int	ly
int	lu
int	Nc
int	k_0
Eigen::VectorXd	theta_0
int	filtorder
rcacFilt	FILT
std::deque< Eigen::VectorXd >	uBar
std::deque< Eigen::VectorXd >	ufBar
std::deque< Eigen::VectorXd >	zBar
std::deque< Eigen::VectorXd >	zfBar
std::deque< Eigen::MatrixXd >	PhiBar
std::deque< Eigen::MatrixXd >	PhifBar
Eigen::MatrixXd	P
Eigen::VectorXd	theta
Eigen::VectorXd	uOut
Eigen::VectorXd	uIn
Eigen::MatrixXd	Phi
Eigen::VectorXd	uphi
Eigen::VectorXd	yphi
bool	rcacRLS
bool	rcacGrad
int	kk = 1

Detailed Description

The parent RCAC class. This class handles all the low level computation of RCAC such as the filtering, coefficient updates, and keeping track of the regressors.

Almost all the methods are polymorphic and can be modified by child classes to create RCAC algorithms with more complex filtering.

Member Function Documentation

coeffUpdate()

virtual void RCAC::coeffUpdate ( Eigen::VectorXd &  zIn )

protectedpure virtual

Abstract function for the ceofficient update implementation (e.g. RLS, Gradient). Child classes must implement this function for RCAC to work

Parameters

zIn	performance measurement

computeFiltered()

void RCAC::computeFiltered ( )

protected

Compute the filtered variables given current regressors and filter values

getCoeff()

Eigen::VectorXd RCAC::getCoeff ( )

inline

Returns a vector of the current RCAC coefficients.

getControl()

Eigen::VectorXd RCAC::getControl ( )

inline

Returns RCAC's computed value for the control. Must run oneStep at least once.

getkk()

int RCAC::getkk ( )

inline

Returns the timestep of RCAC

getlu()

int RCAC::getlu ( )

inline

Returns the number of control inputs that RCAC is using.

getly()

int RCAC::getly ( )

inline

Returns the number of sensor measurements that RCAC is using.

getlz()

int RCAC::getlz ( )

inline

Returns the number of performance measurements that RCAC is using.

getNc()

int RCAC::getNc ( )

inline

Returns the controller order of RCAC.

init()

template<typename T >

RCAC * RCAC::init ( T &  FLAGS, rcacFilt &  FILT, std::string &  rcacType  )

static

Factory method for initializing RCAC types. This method is defined in the file RCACCreator.hpp

Parameters

FLAGS	template for a struct containing the required flags for the type of RCAC to be used.
FILT	struct containing the filter coefficients for ( ).
whichRCAC	string containing the type of RCAC to be used, defined in RCACCreator.hpp.

A factory method that creates a pointer to the specific type of RCAC you want to use.

Parameters

FLAGS	template for a struct containing the required flags for the type of RCAC to be used.
FILT	struct containing the filter coefficients for ( ).
whichRCAC	string containing the type of RCAC to be used.

initFiltered()

void RCAC::initFiltered ( )

protected

Initialize the variables required for filtering to zero

initRegressor()

void RCAC::initRegressor ( )

protected

Initialize the regressor variables uphi and yphi to zero vectors

oneStep()

void RCAC::oneStep	(	Eigen::VectorXd &	uIn,
		Eigen::VectorXd &	zIn,
		Eigen::VectorXd &	yIn
	)

This function allows the user to compute one step of the RCAC update.

Parameters

uIn	previous control input
zIn	current performance measurement
yIn	current sensor measurement

---

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

• RCAC.hpp
• RCAC.cpp
• RCACCreator.hpp