AMCLLocalizer Class Reference

A localization method implementing a simplified AMCL (Adaptive Monte Carlo Localization) approach. More...

#include <AMCLLocalizer.hpp>

Inheritance diagram for AMCLLocalizer:

Collaboration diagram for AMCLLocalizer:

Public Member Functions
	AMCLLocalizer ()
	Default constructor.
nav_msgs::msg::Odometry	get_pose ()
	Gets the current estimated pose as an Odometry message.
tf2::Transform	getEstimatedPose () const
	Gets the current estimated pose as a transform.
virtual void	on_initialize () override
	Initializes the localization method.
void	update (NavState &nav_state) override
	General update of the localization state.
void	update_rt (NavState &nav_state) override
	Real-time update of the localization state.
	~AMCLLocalizer ()
	Destructor.

Protected Member Functions
void	correct (NavState &nav_state)
	Applies the sensor model to update particle weights.
void	initializeParticles ()
	Initializes the set of particles.
void	odom_callback (nav_msgs::msg::Odometry::UniquePtr msg)
	Callback for receiving odometry updates.
void	predict (NavState &nav_state)
	Applies the motion model to update particle poses.
void	publishEstimatedPose (const tf2::Transform &est_pose)
	Publishes the estimated pose with covariance.
void	publishParticles ()
	Publishes the current set of particles.
void	publishTF (const tf2::Transform &map2bf)
	Publishes a TF transform between map and base footprint.
void	reseed ()
	Re-initializes the particle cloud if necessary.
void	update_odom_from_tf ()
	Update odom from TFs instead of a odom topic.

Protected Attributes
std::shared_ptr< Bonxai::ProbabilisticMap >	bonxai_map_
	Internal static map.
bool	compute_odom_from_tf_ {false}
	Whether to use TFs to compute odom.
std::size_t	correct_max_points_ {1500}
double	downsampled_cloud_size_ {0.05}
rclcpp::Publisher< geometry_msgs::msg::PoseWithCovarianceStamped >::SharedPtr	estimate_pub_
	Publisher for the estimated robot pose with covariance.
double	inflation_prob_min_ {0.01}
double	inflation_stddev_ {1.5}
bool	initialized_odom_ = false
	Flag indicating if the odometry has been initialized.
::navmap::NavCelId	last_cid_ {0}
rclcpp::Time	last_input_time_
	Timestamp of the last input message (odometry or initial pose).
tf2::Transform	last_odom_ {tf2::Transform::getIdentity()}
	Previous odometry transform (used to compute deltas).
rclcpp::Time	last_reseed_
	Timestamp of the last reseed event.
double	min_noise_xy_ {0.05}
	Minimum translation noise threshold.
double	min_noise_yaw_ {0.05}
	Minimum yaw noise threshold.
double	noise_rotation_ {0.01}
	Rotational noise standard deviation.
double	noise_translation_ {0.01}
	Translational noise standard deviation.
double	noise_translation_to_rotation_ {0.01}
	Coupling noise between translation and rotation.
tf2::Transform	odom_ {tf2::Transform::getIdentity()}
	Last odometry transform received.
rclcpp::Subscription< nav_msgs::msg::Odometry >::SharedPtr	odom_sub_
	Subscriber for odometry messages.
std::vector< Particle >	particles_
	List of particles representing the belief distribution.
rclcpp::Publisher< geometry_msgs::msg::PoseArray >::SharedPtr	particles_pub_
	Publisher for visualization of the particle cloud.
tf2::Transform	pose_
	Current estimated odometry-based pose.
double	reseed_time_
	Time interval (in seconds) after which the particles should be reseeded.
std::mt19937	rng_
	Random number generator used for sampling noise.
std::unique_ptr< tf2_ros::TransformBroadcaster >	tf_broadcaster_
	TF broadcaster to publish map to base_footprint transform.
double	top_keep_fraction_ {0.2}
double	weights_tau_ {0.7}

Detailed Description

A localization method implementing a simplified AMCL (Adaptive Monte Carlo Localization) approach.

Constructor & Destructor Documentation

AMCLLocalizer()

AMCLLocalizer

(

)

Default constructor.

~AMCLLocalizer()

~AMCLLocalizer ( )

default

Destructor.

Member Function Documentation

correct()

void correct ( NavState & nav_state )

protected

Applies the sensor model to update particle weights.

Parameters

nav_state

The current navigation state.

get_pose()

nav_msgs::msg::Odometry get_pose

(

)

Gets the current estimated pose as an Odometry message.

Returns

A nav_msgs::msg::Odometry message containing the estimated pose.

getEstimatedPose()

tf2::Transform getEstimatedPose

(

)

const

Gets the current estimated pose as a transform.

Returns

The transform from map to base footprint frame.

initializeParticles()

void initializeParticles ( )

protected

Initializes the set of particles.

odom_callback()

void odom_callback ( nav_msgs::msg::Odometry::UniquePtr msg )

protected

Callback for receiving odometry updates.

Parameters

msg	The incoming odometry message.

on_initialize()

void on_initialize ( )

overridevirtual

Initializes the localization method.

Sets up publishers, subscribers, and prepares the particle filter.

Exceptions

std::runtime_error

if initialization fails.

predict()

void predict ( NavState & nav_state )

protected

Applies the motion model to update particle poses.

Parameters

nav_state

The current navigation state.

publishEstimatedPose()

void publishEstimatedPose ( const tf2::Transform & est_pose )

protected

Publishes the estimated pose with covariance.

Parameters

est_pose

The estimated transform to be published.

publishParticles()

void publishParticles ( )

protected

Publishes the current set of particles.

publishTF()

void publishTF ( const tf2::Transform & map2bf )

protected

Publishes a TF transform between map and base footprint.

Parameters

map2bf

The transform to be published.

reseed()

void reseed ( )

protected

Re-initializes the particle cloud if necessary.

update()

void update ( NavState & nav_state )

override

General update of the localization state.

May include operations not suitable for real-time execution.

Parameters

nav_state

The current navigation state (read/write).

update_odom_from_tf()

void update_odom_from_tf ( )

protected

Update odom from TFs instead of a odom topic.

update_rt()

void update_rt ( NavState & nav_state )

override

Real-time update of the localization state.

Used for time-critical update operations.

Parameters

nav_state

The current navigation state (read/write).

Member Data Documentation

bonxai_map_

std::shared_ptr<Bonxai::ProbabilisticMap> bonxai_map_

protected

Internal static map.

compute_odom_from_tf_

bool compute_odom_from_tf_ {false}

protected

Whether to use TFs to compute odom.

correct_max_points_

std::size_t correct_max_points_ {1500}

protected

downsampled_cloud_size_

double downsampled_cloud_size_ {0.05}

protected

estimate_pub_

rclcpp::Publisher<geometry_msgs::msg::PoseWithCovarianceStamped>::SharedPtr estimate_pub_

protected

Publisher for the estimated robot pose with covariance.

inflation_prob_min_

double inflation_prob_min_ {0.01}

protected

inflation_stddev_

double inflation_stddev_ {1.5}

protected

initialized_odom_

bool initialized_odom_ = false

protected

Flag indicating if the odometry has been initialized.

last_cid_

::navmap::NavCelId last_cid_ {0}

protected

last_input_time_

rclcpp::Time last_input_time_

protected

Timestamp of the last input message (odometry or initial pose).

last_odom_

tf2::Transform last_odom_ {tf2::Transform::getIdentity()}

protected

Previous odometry transform (used to compute deltas).

last_reseed_

rclcpp::Time last_reseed_

protected

Timestamp of the last reseed event.

min_noise_xy_

double min_noise_xy_ {0.05}

protected

Minimum translation noise threshold.

min_noise_yaw_

double min_noise_yaw_ {0.05}

protected

Minimum yaw noise threshold.

noise_rotation_

double noise_rotation_ {0.01}

protected

Rotational noise standard deviation.

noise_translation_

double noise_translation_ {0.01}

protected

Translational noise standard deviation.

noise_translation_to_rotation_

double noise_translation_to_rotation_ {0.01}

protected

Coupling noise between translation and rotation.

odom_

tf2::Transform odom_ {tf2::Transform::getIdentity()}

protected

Last odometry transform received.

odom_sub_

rclcpp::Subscription<nav_msgs::msg::Odometry>::SharedPtr odom_sub_

protected

Subscriber for odometry messages.

particles_

std::vector<Particle> particles_

protected

List of particles representing the belief distribution.

particles_pub_

rclcpp::Publisher<geometry_msgs::msg::PoseArray>::SharedPtr particles_pub_

protected

Publisher for visualization of the particle cloud.

pose_

tf2::Transform pose_

protected

Current estimated odometry-based pose.

reseed_time_

double reseed_time_

protected

Time interval (in seconds) after which the particles should be reseeded.

rng_

std::mt19937 rng_

protected

Random number generator used for sampling noise.

tf_broadcaster_

std::unique_ptr<tf2_ros::TransformBroadcaster> tf_broadcaster_

protected

TF broadcaster to publish map to base_footprint transform.

top_keep_fraction_

double top_keep_fraction_ {0.2}

protected

weights_tau_

double weights_tau_ {0.7}

protected

---

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

• localizers/easynav_navmap_localizer/include/easynav_navmap_localizer/AMCLLocalizer.hpp
• localizers/easynav_navmap_localizer/src/easynav_navmap_localizer/AMCLLocalizer.cpp