flztiii / traffic_flow_trajectory_planning

Adaptive Spatio-Temporal Voxels Based Trajectory Planning for Autonomous Driving in Highway Traffic Flow

This repo is the code for the paper "Adaptive Spatio-Temporal Voxels Based Trajectory Planning for Autonomous Driving in Highway Traffic Flow".

demo_s.mp4

Requirements

For running code:

• Ubuntu20.04

• python3.9

• ros-noetic (http://wiki.ros.org/ROS/Installation)

• rospkg (pip install rospkg)

• gym (pip install gym)

• pygame (pip install pygame)

• pcl_ros (sudo apt install ros-noetic-pcl-ros)

• ooqp

• glog

• tk_spline

• odeint-v2

Install

1. create a root directory

mkdir catkin_ws
cd catkin_ws
mkdir src
cd src
catkin_init_workspace

2. clone the code

git clone https://github.com/flztiii/traffic_flow_trajectory_planning.git

3. install

catkin_make -DCMAKE_BUILD_TYPE=Release

How to use

1. start roscore

roscore

2. launch the simulator

cd catkin_ws
source devel/setup.bash
roslaunch simulator start_gym.launch

3. start the planner

cd catkin_ws
source devel/setup.bash
roslaunch local_planning test.launch

press "Enter" key in the terminal to start the program.

Acknowledgement

https://github.com/HKUST-Aerial-Robotics/EPSILON

https://github.com/Farama-Foundation/HighwayEnv