An Open ROS Simulation Framework for TWR-Based Acoustic Localization of Underwater Robots: Design and Experimental Validation

by Fabian Steinmetz, Daniel Duecker, Nils Sichert, Christian Busse, Edwin Kreuzer, Christian Renner

implementation by Nils Sichert

Short installation guide:

Requirements for using the acoustic simulation from csv-files:

• ROS Noetic
• Python 3

git clone https://github.com/nsi18/acoustic_sim.git

additional requirements for using the acoustic simulation in Gazebo, with BlueRov as agent and ability to follow a path:

• QGroundControl

Structur should be: PX4-Autopilot catkin_ws

• build
• devel
• logs
• src
  • bluerov_sim
  • control
  • hippocampus_sim
  • hippocampus_msgs
  • mavlink
  • mavros
  • path_planning
  • acoustic_sim

git clone https://github.com/FormulasAndVehicles/PX4-Autopilot.git

git clone https://github.com/nsi18/acoustic_sim.git
git clone https://github.com/nsi18/bluerov_sim.git
git clone https://github.com/nsi18/control.git
git clone https://github.com/nsi18/hippocampus_common.git
git clone https://github.com/nsi18/hippocampus_msgs.git
git clone https://github.com/nsi18/mavlink.git
git clone https://github.com/nsi18/mavros.git
git clone https://github.com/nsi18/path_planning.git

Don't forget to rebuild your workspace, before running simulation in Gazebo!

Usage

There are 2 independent functionalities:

1. running within Gazebo and doing a live simulation

1. set flag in acoustic_config: "RosRun":true
2. launch simulation with following code:

roslaunch acoustic_sim simulation_path.launch vehicle_type:=bluerov use_external_controller:=true vehicle_name:=bluero

2. simulate from recorded gps-data (must be interpolated to 100Hz!)

1. set flag in acoustic_config: "RosRun":false set GPS-file path in acoustic_sim/src/acoustic_sim/dataloader_class.py
2. run simulation by executing acoustic_sim/nodes/simulation_class.py or:

rosrun acoustic_sim simulation_class.py

Configuration

There are two config-files (acoustic simulation, filter settings): You can finde them: acoustic_sim/config/

Acoustic Simulation settings (/acoustic_config.json)

general settings

• Timings
• SOS: speed of sound
• algortihm: alternating/ broadcast (MAC-protocoll)
• pollcircle: timetrgd (time triggered), lstAcktrgd (last Response and time triggered)
• TimeOut: time soundwave exists

Modem settings

{
    "type": "[agent or anchor]",
    "name": "[Name]",
    "shortname": "[Shortname]",
    "modem": {
               "id": [INT],
               "DelayTime": [Float],
               "PacketReceptionRate": [Float between 1 = 100% and 0 = 0%],
               "dst": "broadcast",
               "packetTyp": "[TYPE_RANGING_POLL or TYPE_RANGING_ACK]"
             },
    "position":[0.8, 2.55,-0.3]        
}

Filter settings (/filter_config.json)

general settings

• filterTyp: UKF/EKF selection of applied filter
• PredictFrequency: frequency of prediction iterations per second
• lengthDatabag: length of a Databag which record the last filter settings of the last few simulations runs; important for correcting the time step
• MeasErrLoc/ MeasErrScale: Mean and standard deviation [m] applied on prediction data

UKF settings

• beta, alpha, kappa, process noise, initial state, inital covariance, covarianz of measurements

EKF settings

• process noise, inital state, inital covariance, covarianz of measurements

Contact

If you have issues with installing the simulator or other questions, feel free to contact:

• Daniel Duecker: [email protected]
• Nils Sichert: [email protected]