StateEstimator Class Reference

Public Member Functions
Interface
void	SetState (const Cfg &_cfg, const std::vector< double > &_uncertainty)
void	ApplyControls (const ControlSet &_controls, const double _dt)
virtual void	ApplyObservations (SensorInterface *const _sensor)=0
const Cfg &	GetEstimatedState () const
	Get the current state estimate.
const std::vector< double > &	GetUncertainty () const
	Get the uncertainty in the estimated state.

Protected Attributes
Internal State
Get the covariance matrix for the state estimate. Todo: Figure out how to represent this.
Robot *const	m_robot
	The owning robot.
Cfg	m_estimatedState
	The current estimated state.
std::vector< double >	m_uncertainty
	Uncertainty in the current estimate.
bool	m_debug {false}
	The debug flag.

Construction
	StateEstimator (Robot *const _robot)
virtual	~StateEstimator ()
static std::unique_ptr< StateEstimator >	Factory (Robot *const _robot, XMLNode &_node)

Detailed Description

Base abstraction for a state estimator object, which combines the controls applied to a hardware robot with sensor data to estimate the robot's true state.

We assume that all process noise/uncertainty is iid and Gaussian-shaped.

Member Function Documentation

Factory()

std::unique_ptr< StateEstimator > StateEstimator::Factory ( Robot *const  _robot, XMLNode &  _node  )

static

Create a dynamically-allocated state estimator from an XML node.

Parameters

_robot	The robot which this state estimator will work for.
_node	The XML node to parse.

Returns

A state estimator of the type specified by _node.

SetState()

void StateEstimator::SetState	(	const Cfg &	_cfg,
		const std::vector< double > &	_uncertainty
	)

Set the current state and uncertainty.

Parameters

_cfg	The state to set.
_uncertainty	The uncertainty in _cfg.

ApplyControls()

void StateEstimator::ApplyControls	(	const ControlSet &	_controls,
		const double	_dt
	)

Apply a set of controls to the current estimated state.

Note

The Actuator class should define the force/velocity uncertainty, if any.

Parameters

_controls	The set of controls to apply.
_dt	The length of time to apply the controls.

ApplyObservations()

virtual void StateEstimator::ApplyObservations ( SensorInterface *const  _sensor )

pure virtual

Apply the most recent observations from a sensor to the current estimated state.

Parameters

_sensor

The sensor to take observations from.

Implemented in AverageEstimator.

---

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

• StateEstimator.h
• StateEstimator.cpp