romea_ros2_robot_to_human_localisation_core

This package provides robot to human localisation node able to estimate the human leader position by a Kalman filter using observations provided by three localisation plugins: odo (gitlab, github), imu (gitlab, github) and robot to human rtls (gitlab, github) plugins . This plugins are used to convert data coming from sensors to observations that can be used by the Kalman filter.

ROS2 core node description

1) Subscribed Topics

• localisation/twist (romea_localisation_msgs::msg::ObservationTwist2DStamped)

  This topic is provided by odo localisation plugin node and contains robot twist displacement observation deduced from odometry data coming from controller

• localisation/angular_speed (romea_localisation_msgs::msg::ObservationAngularSpeedStamped)

  This topic is provided by imu localisation plugin node and contains robot angular speed observation deduced from data coming from IMU sensor

• localisation/position (romea_localisation_msgs::msg::ObservationPosition2DStamped)

  This topic is provided by rtls localisation plugin node and contains rough estimation of the position of the human leader computed by trilateration algorithm base on ranging data between rtls transceivers embedded on the robot and the rtls transceiver carried by human leader

• localisation/range (romea_localisation_msgs::msg::ObservationRangeStamped)

  This topic is provided by rtls localisation plugin node and contains range observation deduced from ranging data between rtls transceivers embedded on the robot and the rtls transceiver carried by human leader

2) Published Topics

• leader_position (romea_common_msgs::msg::Position2DStamped)

  Position of human leader in the robot localisation reference frame

3) Parameters

~filter.state_pool_size (int , default: 1000)

Size of the pool of filter states

~predictor.leader_motion_noise_std (double, default: 1)

Standard deviation on leader position (along each axis) using in the constant position kalman filter motion model in order to predict leader position at each time

~predictor.maximal_dead_recknoning_elapsed_time (double, default: 1.0)

Maximal elapsed time in dead reckoning mode before to stop localisation filter

~predictor.maximal_dead_recknoning_travelled_distance (double, default: 1.0)

Maximal travelled distance in dead reckoning mode before to stop localisation filter

~predictor.maximal_position_circular_error_probability (double, default)

Maximal circular error in dead reckoning mode before to stop localisation filter

~twist_updater.minimal_rate (int, default: 10)

Minimal rate for twist_updater input data (provided by odo plugin), if this rate is equal to 0 twist_updater is not started

~linear_speed_updater.minimal_rate (int, default: 0)

Minimal rate for linear_speed_updater input data (provided by odo plugin), if this rate is equal to 0 linear_speed_updater is not started

~linear_speeds_updater.minimal_rate (int, default: 0)

Minimal rate for linear_speeds_updater input data (provided by odo plugin), if this rate is equal to 0 linear_speeds_updater is not started

~angular_updater.minimal_rate (int, default: 0)

Minimal rate for angular_updater input data (provided by imu plugin), if this rate is equal to 0 angular_updater is not started

~position_updater.mahalanobis_distance_rejection_threshold (double, default: 5.0)

Mahalanobis distance taking into account by position updater to reject outliers

~position_updater.minimal_rate (int, default: 1)

Minimal rate for position_updater input data (provided by rtls plugin), if this rate is equal to 0 position_updater is not started

~position_updater.trigger (string, default: "once")

Update trigger mode when position data is received. If "once" mode is selected the position updater will be triggered only one time otherwise the position updater will be triggerred each time data is received.

~range_updater.mahalanobis_distance_rejection_threshold (double, default: 5.0)

Mahalanobis distance taking into account by range updater to reject outliers

~range_updater.minimal_rate (int, default: 10)

Minimal rate for position_updater input data (provided by rtls plugin); , if this rate is equal to 0 range_updater is not started

~range_updater.trigger (string, default: "always")

Update trigger mode when position data is received. If "once" mode is selected the range updater will be triggered only one time otherwise the range updater will be triggerred each time data is received.

~range_updater.use_constraints (bool)

Option to ensure that the x-axis position is always positive. It's useful when only two rtls transceivers is used to localisation human leader because in this situation it is impossible to know whether the leader is in front or behind the vehicle.

~base_footprint_frame_id (string, default: base_footprint):

Name of robot base footprint

~publish_rate (int, default: 10)

Rate at which localisation results are published

~log_directory (string, default: result of rclcpp::get_logging_directory()):

Directory where localisation logs are stored

~debug (bool, default: false)

Enable debug logs

Usage

See romea_localisation_bringup project

Contributing

If you'd like to contribute to this project, here are some guidelines:

1. Fork the repository.
2. Create a new branch for your changes.
3. Make your changes.
4. Write tests to cover your changes.
5. Run the tests to ensure they pass.
6. Commit your changes.
7. Push your changes to yo

License

This project is released under the Apache License 2.0. See the LICENSE file for details.

Authors

romea_robot_to_human_localisation_core project was developed by Jean Laneurit in the context of BaudetRob2 ANR project.

Contact

If you have any questions or comments about romea_robot_to_human_localisation_core project, please contact Jean Laneurit