mppi_swerve_drive_ros

MPPI (Model Predictive Path-Integral) Controller for a Swerve Drive Robot

[Website] [PDF] [Arxiv] [IEEE Xplore] [Poster]

iros2024_movie_20240912.mp4

Citation

If you use this work in an academic context, please cite the following publication:

@inproceedings{mizuho2024iros
  author={Aoki, Mizuho and Honda, Kohei and Okuda, Hiroyuki and Suzuki, Tatsuya},
  booktitle={2024 IEEE/RSJ International Conference on Intelligent Robots and Systems (IROS)}, 
  title={Switching Sampling Space of Model Predictive Path-Integral Controller to Balance Efficiency and Safety in 4WIDS Vehicle Navigation}, 
  year={2024},
  volume={},
  number={},
  pages={3196-3203},
  doi={10.1109/IROS58592.2024.10802359}}

Setup

[Option 1] Docker environment

CLICK HERE TO EXPAND

1. Prerequisites

   • docker
     • For ubuntu users:

       curl -fsSL https://get.docker.com -o get-docker.sh
       sudo sh get-docker.sh
       

   • rocker
     • For ubuntu users:

       sudo apt-get install python3-rocker
       

2. Clone the project repository.

   cd <path to your workspace>
   git clone https://github.com/MizuhoAOKI/mppi_swerve_drive_ros
   

3. Run for the first time setup to build the docker image.

   cd <path to your workspace>/mppi_swerve_drive_ros
   make setup_docker
   

4. Launch the docker container and get into the bash inside.

   cd <path to your workspace>/mppi_swerve_drive_ros
   make run_docker
   

5. [Inside the docker container] Build the project.

   cd ~/mppi_swerve_drive_ros
   make build
   

[Option 2] Native environment

CLICK HERE TO EXPAND

1. Prerequisites

   • ubuntu 20.04
   • ros noetic

2. Clone the project repository.

   cd <path to your workspace>
   git clone https://github.com/MizuhoAOKI/mppi_swerve_drive_ros
   

3. Install foundation packages.

   cd <path to your workspace>/mppi_swerve_drive_ros
   sudo make install_deps
   

4. Initialize rosdep, update it, and install dependencies.

   cd <path to your workspace>/mppi_swerve_drive_ros
   sudo rosdep init
   rosdep update
   rosdep update && rosdep install -y --from-paths src --ignore-src --rosdistro noetic
   

5. Build the project.

   cd <path to your workspace>/mppi_swerve_drive_ros
   make build
   

Build

Build the project.

cd <path to your workspace>/mppi_swerve_drive_ros
make build

(Optional) Clean the cache before building the project if necessary.

cd <path to your workspace>/mppi_swerve_drive_ros
make clean

Usage

[Case 1] Launch gazebo simulator only, operating a 4wids vehicle manually with a joypad.

cd <path to your workspace>/mppi_swerve_drive_ros
source /opt/ros/noetic/setup.bash && source ./devel/setup.bash
roslaunch launch/gazebo_world.launch gazebo_world_name:=maze

• gazebo_world_name options:
  • empty
  • empty_garden
  • cylinder_garden
  • maze
• Default joystick path is /dev/input/js0. If you want to change the path, please edit mppi_swerve_drive_ros/src/operation/joy_controller/config/joy.yaml.

[Case 2] Navigate a 4wids vehicle autonomously with a MPPI controller.

• Try MPPI-3D(a) (driving faster but dangerous sometimes)

  cd <path to your workspace>/mppi_swerve_drive_ros
  source /opt/ros/noetic/setup.bash && source ./devel/setup.bash
  roslaunch launch/navigation.launch local_planner:=mppi_3d_a

• Try MPPI-3D(b) (relatively safe but driving slower)

  cd <path to your workspace>/mppi_swerve_drive_ros
  source /opt/ros/noetic/setup.bash && source ./devel/setup.bash
  roslaunch launch/navigation.launch local_planner:=mppi_3d_b

• Try MPPI-4D (safe but relatively slow)

  cd <path to your workspace>/mppi_swerve_drive_ros
  source /opt/ros/noetic/setup.bash && source ./devel/setup.bash
  roslaunch launch/navigation.launch local_planner:=mppi_4d

• [Author's Recommendation] ⭐Try MPPI-H⭐ (good balance between quickness and safety)

  cd <path to your workspace>/mppi_swerve_drive_ros
  source /opt/ros/noetic/setup.bash && source ./devel/setup.bash
  roslaunch launch/navigation.launch local_planner:=mppi_h

demo_mppi_swerve_drive_ros.mp4

Note

Due to asynchronous simulation on ROS and the sampling-based algorithm relying on multi-threading computation, the controllers' performance can vary depending on a user's environment.