GROMI SCAN PLANNING

This is based on the jackal velodyne simulation package.

Introduction

Objective

The goal is to autonomously navigate and map an unknown area using SLAM and a frontier exploration algorithm based on laser data from a Velodyne VLP-16 LIDAR.

How it works

It is demostrated on the simulated environment. First, it tries to find the frontier locations which is the border of the free space and unknown space. It then decides the next goal location among the frontier locations and calculates the best scan location near the next goal location. For your information, in the videos, the blue marks are the frontier locations and the red circles are the designated scan locations.

Installation

Tested on Ubuntu 18.04 with ROS Melodic
To run this code, please install the following packages:

sudo apt-get install ros-melodic-navigation
sudo apt-get install ros-melodic-tf2-sensor-msgs
sudo apt-get install ros-melodic-jackal-*
sudo apt-get install ros-melodic-velodyne-*
sudo apt-get install ros-melodic-octomap-mapping 
sudo apt-get install ros-melodic-pointcloud-to-laserscan
sudo apt-get install python-scipy

python -m pip install --upgrade pip
pip install scikit-build
pip install cmake
pip install opencv-python

How to run

simulation with gmapping : roslaunch gromi_scan_planning simulation_gromi_gmaaping.launch
simulation with lego-loam : roslaunch gromi_scan_planning simulation_gromi.launch
real gromi hardware : roslaunch gromi_scan_planning gromi_scanning.launch

Nodes

Exploration Node

exlore.py

This node is the brain that computes what the next goal position and scan location should be the robot should go to.

Parameters

self.minScanDist = 5 : no duplicate scan within 5 m
self.min_bndX = -10 : Target area X min bound
self.max_bndX = 10 : Target area X max bound
self.min_bndY = -10 : Target area Y min bound
self.max_bndY = 10 : Target area Y max bound

Subscribed Topics:

/map for updating the current map
/move_base/global_costmap/costmap for updating the current global costmap
/move_base/global_costmap/costmap_updates for updating the current global costmap
/move_base/result for determining when the robot has reached its goal position
/move_base/feedback for determining where the robot is in the map at any given point in time
/scan_cloud for subscribing the front 2d Lidar point cloud data
/octomap_point_cloud_centers for subscribing octomap point cloud center data
/finish_scan for subscribing to stop the static scanning process

Published Topic:

/move_base_simple/goal for the next goal location
/visualization_marker for visualizing the frontier locations (blue markers)
/scan_locations for visualizing the scan locations (red circles)
/start_scan for starting the static scanning process

gromi_static_scan.py

This node is to take pictures and run/stop the rotation scanning motor.

Parameters

self.rotation_time = 100 : scanning time for one rotation
self.num_of_pic = 8 : number of pictures that take for one rotation
self.time_between_pics : time between each picture
self.trigger : trigger to take pictures and run/stop the rotation scanning motor
self.cnt : number of pictures that take at current

Subscribed Topics:

/start_scan for receiving command that start the static scanning process

Published Topic:

/run_scan_motor for sending command that run/stop the rotation scanning motor
/finish_scan for sending the command that finish the static scanning process

motor_controller.py

This node is to make differential wheel speed.

Parameters

self.mps_2_data_coeff = 212.6 : Conversion factor for converting m/s to a data value for the motor controller. Setting this to a higher value will lead to the robot moving faster for the same speed input.
self.radps_2_data_coeff = 233.9 : Conversion factor for converting rad/s to a data value offset for the motor controller. Setting this to a higher value will lead to the robot rotating faster for the same angular speed input.
self.smoothing_coeff = 0.8 : The smoothing coefficient controls how much smoothing is applied. Must be between [0, 1). Setting to 0.0 would apply zero smoothing while setting to 1.0 would make the system unresponsive. Keep in mind that higher publishing frequency requires a higher smoothing coefficient for the same result
self.pub_frequency = 20.0 : The controller's running frequency in Hz

Subscribed Topics:

/cmd_vel/nav for receiving velocity command

Published Topic:

/left_wheel_vel for sending the left wheel velocity command
/right_wheel_vel for sending the right wheel velocity command

scan_motor_controller.py

This node is to make rotation of scanning system.

Parameters

self.start_command = "0106007D40002812" : define the start commands for the scanning motor
self.stop_command = "0106007D0020180A" : define the stop commands for the scanning motor
self.run_motor : true - run the scanning motor, false - stop the scanning motor
self.pub_frequency = 20.0 : The controller's running frequency in Hz

Subscribed Topics:

/run_scan_motor for receiving command that start to run the scanning motor

Demo Video

A video of Jackal exploring a simulated environment (rectangular objects) in Gazebo and visualizing in RViZ. here.
A video of Jackal exploring a simulated environment (cylinder objects) in Gazebo and visualizing in RViZ. here.

kimfeel / gromi_scan_planning Goto Github PK