The Odometry class handles odometry readings (2D pose and velocity with related timestamp) More...

#include <odometry.h>

Public Types
typedef boost::function< void(double, double)>	IntegrationFunction
	Integration function, used to integrate the odometry:
Public Member Functions
double	getAngularEstimated () const
	Retrieves the angular velocity estimation.
double	getHeading () const
	heading getter
double	getLinearEstimated () const
	Retrieves the linear velocity estimation.
ros::Time	getTimestamp () const
	timestamp getter
double	getX () const
	x position getter
double	getY () const
	y position getter
	Odometry (size_t velocity_rolling_window_size=10)
	Constructor Timestamp will get the current time value Value will be set to zero.
void	setWheelParams (double wheel_separation, double wheel_radius)
	Sets the wheel parameters: radius and separation.
bool	update (double left_pos, double right_pos, const ros::Time &time)
	Updates the odometry class with latest wheels position.
Private Types
typedef bacc::accumulator_set < double, bacc::stats < bacc::tag::rolling_mean > >	RollingMeanAcc
	Rolling mean accumulator and window:
typedef bacc::tag::rolling_window	RollingWindow
Private Member Functions
void	integrateExact (double linear, double angular)
	Integrates the velocities (linear and angular) using exact method.
void	integrateRungeKutta2 (double linear, double angular)
	Integrates the velocities (linear and angular) using 2nd order Runge-Kutta.
Private Attributes
RollingMeanAcc	angular_acc_
double	heading_
IntegrationFunction	integrate_fun_
	Integration funcion, used to integrate the odometry:
double	left_wheel_old_pos_
	Previou wheel position/state [rad]:
RollingMeanAcc	linear_acc_
	Rolling meand accumulators for the linar and angular velocities:
double	right_wheel_old_pos_
ros::Time	timestamp_
	Current timestamp:
double	wheel_radius_
double	wheel_separation_
	Wheel kinematic parameters [m]:
double	x_
	Current pose:
double	y_

Detailed Description

The Odometry class handles odometry readings (2D pose and velocity with related timestamp)

Definition at line 57 of file odometry.h.

Member Typedef Documentation

typedef boost::function<void(double, double)> diff_drive_controller::Odometry::IntegrationFunction

Integration function, used to integrate the odometry:

Definition at line 62 of file odometry.h.

typedef bacc::accumulator_set<double, bacc::stats<bacc::tag::rolling_mean> > diff_drive_controller::Odometry::RollingMeanAcc [private]

Rolling mean accumulator and window:

Definition at line 139 of file odometry.h.

typedef bacc::tag::rolling_window diff_drive_controller::Odometry::RollingWindow [private]

Definition at line 140 of file odometry.h.

Constructor & Destructor Documentation

diff_drive_controller::Odometry::Odometry ( size_t velocity_rolling_window_size = 10 )

Constructor Timestamp will get the current time value Value will be set to zero.

Parameters:

velocity_rolling_window_size Rolling window size used to compute the velocity mean

Definition at line 48 of file odometry.cpp.

Member Function Documentation

double diff_drive_controller::Odometry::getAngularEstimated ( ) const

Retrieves the angular velocity estimation.

Returns:: Rolling mean estimation of the angular velocity [rad/s]

Definition at line 103 of file odometry.cpp.

double diff_drive_controller::Odometry::getHeading ( ) const [inline]

heading getter

Returns:: heading [rad]

Definition at line 85 of file odometry.h.

double diff_drive_controller::Odometry::getLinearEstimated ( ) const

Retrieves the linear velocity estimation.

Returns:: Rolling mean estimation of the linear velocity [m]

Definition at line 98 of file odometry.cpp.

ros::Time diff_drive_controller::Odometry::getTimestamp ( ) const [inline]

timestamp getter

Returns:: timestamp

Definition at line 112 of file odometry.h.

double diff_drive_controller::Odometry::getX ( ) const [inline]

x position getter

Returns:: x position [m]

Definition at line 94 of file odometry.h.

double diff_drive_controller::Odometry::getY ( ) const [inline]

y position getter

Returns:: y positioin [m]

Definition at line 103 of file odometry.h.

void diff_drive_controller::Odometry::integrateExact	(	double	linear,
		double	angular
	)		`[private]`

Integrates the velocities (linear and angular) using exact method.

Other possible integration method provided by the class.

Parameters:

linear	Linear velocity [m] (linear displacement, i.e. m/s * dt) computed by encoders
angular	Angular velocity [rad] (angular displacement, i.e. m/s * dt) computed by encoders
linear
angular

Exact integration (should solve problems when angular is zero):

Definition at line 132 of file odometry.cpp.

void diff_drive_controller::Odometry::integrateRungeKutta2	(	double	linear,
		double	angular
	)		`[private]`

Integrates the velocities (linear and angular) using 2nd order Runge-Kutta.

Parameters:

linear	Linear velocity [m] (linear displacement, i.e. m/s * dt) computed by encoders
angular	Angular velocity [rad] (angular displacement, i.e. m/s * dt) computed by encoders

Runge-Kutta 2nd order integration:

Normalization of angle to [-Pi, Pi]:

Definition at line 114 of file odometry.cpp.