Skip to content

a Robust and Efficiency navigation framework for indoor wheel-based robot

Notifications You must be signed in to change notification settings

Lkaho/Online-Motion-Planning-based-on-Bezier

Repository files navigation

A Robust and Efficent Navigation Framework for Indoor Wheel-based Robot

In this project, we designed a robust and efficent navigation framework for indoor mobile robot in unknown environments. For the motion planning part, based on the onboard state estimation and environment perception, we implemented A* for front-end path searching and generated back-end corridor-based trajectory by using Bezier curves. For the control part, we implemented path tracking controller by using LQR and MPC algorithm respectively.

1. Main contribution

  • We implement A-star method for front-end path searching.(see navigation/src/astar.cpp)
  • We implement corridor generator for extracting safe region from occupancy map for mobile robot. (see navigation/src/safe_corridor_generator.cpp)
  • We ues OSQP_Interface to solve QP problem to generate Bezier curve. Thanks to the Hodograph property of Bezier curves,, we are able to bound the position within safe regions and limit the higher order dynamics of the trajectory within feasible regions.(see navigation/src/trajectory_generator.cpp)
  • We implement LQR path tracking controller to track the path.(see controller/src)

2. Installation

Dependencies

This software is built on the Robotic Operating System (ROS), which needs to be installed first. Additionally, the Navigation Framework depends on following software:

Building

In order to install the Navigation Framework, clone the latest version from this repository into your catkin workspace and compile the package using ROS.

cd catkin_workspace/src
git clone https://github.com/Lkaho/Online-Motion-Planning-based-on-Bezier.git
cd Online-Motion-Planning-based-on-Bezier
git submodule init
git submodule update
cd ../../
catkin config --cmake-args -DCMAKE_BUILD_TYPE=Release
catkin build

Basic Usage

TurtleBot3 Waffle Simulation

A running example is provided, making use of the Turtlebot3 simulation environment.

To start with, the Turtlebot3 simulation dependencies need to be installed:

sudo apt install ros-noetic-turtlebot3*

Launch the turtlebot3 gazebo simulation:

export TURTLEBOT3_MODEL=waffle
roslaunch turtlebot3_gazebo turtlebot3_house.launch

Run turtlebot3 slam for visulization:

export TURTLEBOT3_MODEL=waffle
roslaunch turtlebot3_slam turtlebot3_slam.launch slam_methods:=gmapping

Run navigation system:

cd your_catkin_workspace/
source devel/setup.bash
roslaunch navigation start_sim.launch

Then, the launch file will launch Bezier trajectory generation, LQR controler modules.

Click '2D goal' on the toolbar to set the destination to robot.

3. Simulation demo

img

4. Real World Experiment

  • img

  • img

About

a Robust and Efficiency navigation framework for indoor wheel-based robot

Resources

Stars

Watchers

Forks

Releases

No releases published

Packages

No packages published