mars::BindSensorData< T > Class Template Reference

The BaseSensorData class binds the sensor state and covariance matrix. More...

#include </home/runner/work/mars_lib/mars_lib/source/mars/include/mars/sensors/bind_sensor_data.h>

Collaboration diagram for mars::BindSensorData< T >:

Public Member Functions
	BindSensorData ()
void	set_cov (const Eigen::MatrixXd &cov)
	set_cov Takes a full covariance and separates sensor covariance and sensor-core cross-correlation More...
Eigen::MatrixXd	get_full_cov () const
	get_full_cov builds the full covariance matrix More...

Public Attributes
EIGEN_MAKE_ALIGNED_OPERATOR_NEW T	state_
int	full_cov_size_
	size of the full covariance More...
Eigen::MatrixXd	sensor_cov_
	covariance of the sensor states More...
Eigen::MatrixXd	core_sensor_cross_cov_
	cross-correlation between sensor states and the core More...

Detailed Description

template<typename T>
class mars::BindSensorData< T >

The BaseSensorData class binds the sensor state and covariance matrix.

The state is instanciated with the passed template (sensor state type) The sensor state class needs to define the error state. The BaseSensorData class initializes the covariance matrix based on this value.

Constructor & Destructor Documentation

BindSensorData()

template<typename T >

mars::BindSensorData< T >::BindSensorData ( )

inline

  {
    core_sensor_cross_cov_ = Eigen::MatrixXd::Zero(CoreStateType::size_error_, state_.cov_size_);
    sensor_cov_ = Eigen::MatrixXd::Zero(state_.cov_size_, state_.cov_size_);
    full_cov_size_ = CoreStateType::size_error_ + state_.cov_size_;
  }

Member Function Documentation

set_cov()

template<typename T >

void mars::BindSensorData< T >::set_cov ( const Eigen::MatrixXd &  cov )

inline

set_cov Takes a full covariance and separates sensor covariance and sensor-core cross-correlation

Parameters

cov

  {
    // TODO(chb) allow this with changing sensor state sizes
    sensor_cov_ = cov.block(CoreStateType::size_error_, CoreStateType::size_error_, state_.cov_size_, state_.cov_size_);
    core_sensor_cross_cov_ = cov.block(0, CoreStateType::size_error_, CoreStateType::size_error_, state_.cov_size_);
  }

get_full_cov()

template<typename T >

Eigen::MatrixXd mars::BindSensorData< T >::get_full_cov ( ) const

inline

get_full_cov builds the full covariance matrix

Returns

sensor covariance and cross covariance, Core entrys are zero.

  {
    // TODO(chb) allow this with changing sensor state sizes
    Eigen::MatrixXd full_cov;
    full_cov.resize(full_cov_size_, full_cov_size_);
 
    // Set core elements to zero
    full_cov.block(0, 0, CoreStateType::size_error_, CoreStateType::size_error_) =
        Eigen::MatrixXd::Zero(CoreStateType::size_error_, CoreStateType::size_error_);
 
    // Fill sensor covariance
    full_cov.block(CoreStateType::size_error_, CoreStateType::size_error_, state_.cov_size_, state_.cov_size_) =
        sensor_cov_;
 
    // Fill cross covariance
    full_cov.block(0, CoreStateType::size_error_, CoreStateType::size_error_, state_.cov_size_) =
        core_sensor_cross_cov_;
    full_cov.block(CoreStateType::size_error_, 0, state_.cov_size_, CoreStateType::size_error_) =
        core_sensor_cross_cov_.transpose();
 
    return full_cov;
  }

Here is the caller graph for this function:

Member Data Documentation

state_

template<typename T >

EIGEN_MAKE_ALIGNED_OPERATOR_NEW T mars::BindSensorData< T >::state_

full_cov_size_

template<typename T >

int mars::BindSensorData< T >::full_cov_size_

size of the full covariance

sensor_cov_

template<typename T >

Eigen::MatrixXd mars::BindSensorData< T >::sensor_cov_

covariance of the sensor states

core_sensor_cross_cov_

template<typename T >

Eigen::MatrixXd mars::BindSensorData< T >::core_sensor_cross_cov_

cross-correlation between sensor states and the core

---

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

• bind_sensor_data.h