KalmanProcessModelBase Class Reference

Virtual base class defining the functions necessary to implement a KalmanBallLocator process model. More...

#include <BaseModel.h>

Inheritance diagram for KalmanProcessModelBase:

[legend]List of all members.

Public Member Functions
	KalmanProcessModelBase ()
virtual	~KalmanProcessModelBase ()
virtual const char *	getModelName () const=0
	Returns the name of the process model.
virtual int	getStateDim () const=0
	Returns the dimension of the state.
virtual void	getP (double *pDest) const=0
	Retrieves the internal kalman filter covariance matrix.
virtual void	getQ (double *pDest) const=0
	Retrieves the process error covariance matrix.
virtual void	setQ (const double *pSource)=0
	Sets the process error covariance matrix.
virtual void	getR (double *pDest) const=0
	Retrieves the measurement error covariance matrix.
virtual void	setR (double *pSource)=0
	Sets the measurement error covariance matrix.
virtual void	reset ()=0
	Resets the process model.
virtual KalmanUpdateResult	update (double time, double x, double y, const RobotPose &pose, double panningVelocity)=0
	Performs an update step of the process model.
virtual KalmanPredictResult	predict (double time)=0
	Permorms a prediction of the ball state.
virtual void	adapt (const OdometryData &lastOdometry, const OdometryData &actualOdometry)=0
	Function to adapt the state of the model when position and rotation of the robot changes.
virtual bool	isNonSensePos (double x, double y) const
	Returns true, if the passed ball position is nonsense (because it is far away from the field).
virtual bool	useModel () const
	Returns true, if the model is to be used.
virtual Out &	toStream (Out &stream) const
	Writes data of the actual process model data to a stream (for debugging).
virtual In &	fromStream (In &stream)
	Reads settings for a process model from a stream (e.g.
Protected Member Functions
void	OutputException (const char *pszLoc, MVException &m) const
	Outputs information about the process model that caused an exception and information about the type of exception.
Protected Attributes
bool	bUseModel
	true, if the model is to be used
Static Protected Attributes
double	pi = 3.1415926535897932384626433832795
	constant pi
double	pi2 = 2*3.1415926535897932384626433832795
	constant 2*pi
double	dFieldDiagonalLength
	1.2*length of field diagonal

---

Detailed Description

Virtual base class defining the functions necessary to implement a KalmanBallLocator process model.

Definition at line 65 of file BaseModel.h.

---

Constructor & Destructor Documentation

KalmanProcessModelBase::KalmanProcessModelBase (   )

Definition at line 31 of file BaseModel.cpp.

virtual KalmanProcessModelBase::~KalmanProcessModelBase (   )  [inline, virtual]

Definition at line 69 of file BaseModel.h.

---

Member Function Documentation

virtual const char* KalmanProcessModelBase::getModelName (   )  const [pure virtual]

Returns the name of the process model. Returns: name of process model Implemented in KalmanConstantSpeedModel, and KalmanFixedPositionModel. Referenced by OutputException(), and toStream().

virtual int KalmanProcessModelBase::getStateDim (   )  const [pure virtual]

Returns the dimension of the state. Returns: dimension of the state Implemented in KalmanConstantSpeedModel, and KalmanFixedPositionModel. Referenced by toStream().

virtual void KalmanProcessModelBase::getP (  double *  pDest  )  const [pure virtual]

Retrieves the internal kalman filter covariance matrix. Parameters: pDest  destination to store the covariance matrix to Implemented in KalmanConstantSpeedModel, and KalmanFixedPositionModel.

virtual void KalmanProcessModelBase::getQ (  double *  pDest  )  const [pure virtual]

Retrieves the process error covariance matrix. Parameters: pDest  destination to store the covariance matrix to Implemented in KalmanConstantSpeedModel, and KalmanFixedPositionModel. Referenced by toStream().

virtual void KalmanProcessModelBase::setQ (  const double *  pSource  )  [pure virtual]

Sets the process error covariance matrix. Parameters: pSource  array containing the covariance matrix entries Implemented in KalmanConstantSpeedModel, and KalmanFixedPositionModel. Referenced by fromStream().

virtual void KalmanProcessModelBase::getR (  double *  pDest  )  const [pure virtual]

Retrieves the measurement error covariance matrix. Parameters: pDest  destination to store the covariance matrix to Implemented in KalmanConstantSpeedModel, and KalmanFixedPositionModel. Referenced by toStream().

virtual void KalmanProcessModelBase::setR (  double *  pSource  )  [pure virtual]

Sets the measurement error covariance matrix. Parameters: pSource  array containing the covariance matrix entries Implemented in KalmanConstantSpeedModel, and KalmanFixedPositionModel. Referenced by fromStream().

virtual void KalmanProcessModelBase::reset (   )  [pure virtual]

Resets the process model. Implemented in KalmanConstantSpeedModel, and KalmanFixedPositionModel. Referenced by fromStream().

virtual KalmanUpdateResult KalmanProcessModelBase::update (  double  time, double  x, double  y, const RobotPose &  pose, double  panningVelocity )  [pure virtual]

Performs an update step of the process model. Parameters: time  time in seconds when the percepts were determined x  measured x-position of the ball in meters relative to the robot y  measured y-position of the ball in meters relative to the robot pose  the current robot pose, used for visualization panningVelocity  the head panning angular velocity, used to adjust the R covariance matrix Returns: the update results of the process model Implemented in KalmanConstantSpeedModel.

virtual KalmanPredictResult KalmanProcessModelBase::predict (  double  time  )  [pure virtual]

Permorms a prediction of the ball state. Parameters: time  actual time in seconds Returns: the predict results of the process model Implemented in KalmanConstantSpeedModel, and KalmanFixedPositionModel.

virtual void KalmanProcessModelBase::adapt (  const OdometryData &  lastOdometry, const OdometryData &  actualOdometry )  [pure virtual]

Function to adapt the state of the model when position and rotation of the robot changes. Parameters: lastOdometry  last odometry of robot (original values in millimeters are used) actualOdometry  actual odometry of robot (original values in millimeters are used) Implemented in KalmanConstantSpeedModel, and KalmanFixedPositionModel.

bool KalmanProcessModelBase::isNonSensePos (  double  x, double  y )  const [virtual]

Returns true, if the passed ball position is nonsense (because it is far away from the field). Returns: true, if passed position is nonsense Definition at line 36 of file BaseModel.cpp. References dFieldDiagonalLength.

virtual bool KalmanProcessModelBase::useModel (   )  const [inline, virtual]

Returns true, if the model is to be used. Returns: true, if model is to be used Definition at line 145 of file BaseModel.h. References bUseModel.

Out & KalmanProcessModelBase::toStream (  Out &  stream  )  const [virtual]

Writes data of the actual process model data to a stream (for debugging). Parameters: stream  stream to write the process model data to Returns: reference to the passed stream Definition at line 45 of file BaseModel.cpp. References bUseModel, getModelName(), getQ(), getR(), and getStateDim().

Here is the call graph for this function:

In & KalmanProcessModelBase::fromStream (  In &  stream  )  [virtual]

Reads settings for a process model from a stream (e.g. when covariance matrices are changed by user. Parameters: stream  stream to read settings from Returns: reference to passed stream Definition at line 75 of file BaseModel.cpp. References bUseModel, reset(), setQ(), and setR().

Here is the call graph for this function:

void KalmanProcessModelBase::OutputException (  const char *  pszLoc, MVException &  m )  const [protected]

Outputs information about the process model that caused an exception and information about the type of exception. Parameters: pszLoc  description of calculation step the exception occured in m  reference to exception Definition at line 105 of file BaseModel.cpp. References MVException::getDescription(), getModelName(), idText, and OUTPUT. Referenced by KalmanFixedPositionModel::predict(), KalmanConstantSpeedModel::predict(), KalmanFixedPositionModel::update(), and KalmanConstantSpeedModel::update().

Here is the call graph for this function:

---

Member Data Documentation

bool KalmanProcessModelBase::bUseModel [protected]

true, if the model is to be used Definition at line 171 of file BaseModel.h. Referenced by fromStream(), toStream(), and useModel().

double KalmanProcessModelBase::pi = 3.1415926535897932384626433832795 [static, protected]

constant pi Definition at line 13 of file BaseModel.cpp. Referenced by KalmanFixedPositionModel::update(), and KalmanConstantSpeedModel::update().

double KalmanProcessModelBase::pi2 = 2*3.1415926535897932384626433832795 [static, protected]

constant 2*pi Definition at line 14 of file BaseModel.cpp.

double KalmanProcessModelBase::dFieldDiagonalLength [static, protected]

Initial value: 1.2*sqrt((2.0*xPosOpponentGoal)*(2.0*xPosOpponentGoal) + (2.0*yPosLeftFlags)*(2.0*yPosLeftFlags)) 1.2*length of field diagonal Definition at line 15 of file BaseModel.cpp. Referenced by isNonSensePos().

---

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

• Modules/BallLocator/GT2004ProcessModels/BaseModel.h
• Modules/BallLocator/GT2004ProcessModels/BaseModel.cpp

---