Repository files navigation

This repository contains the code for the paper:

Efficient Trajectory Planning for Multiple Non-holonomic Mobile Robots via Prioritized Trajectory Optimization

Authors: Juncheng Li, Maopeng Ran, and Lihua Xie from Nanyang Technological University.

Accepted in IEEE Robotics and Automation Letters (RA-L). You can find the full-text paperhere.

This paper proposes an efficient trajectory planning approach that generates safe, dynamically feasible and near-optimal trajectoriesfor multiple non-holonomic mobile robots in obstacle-rich environments.

Please click in the image to see our video:

• Ubuntu 16.04 to 20.04
• ROS Kinetic
• Octomap
• Ipopt

FollowROS Installation

FollowIpopt Installation

sudo apt-get install ros-kinetic-octomap*sudo apt-get install ros-kinetic-dynamic-edt-3dsudo apt-get install cppad

cd ~/catkin_ws/srcgit clone https://github.com/LIJUNCHENG001/multi_robot_traj_planner.gitcd ../ && catkin_makesource ~/catkin_ws/devel/setup.bash

roslaunch multi_robot_traj_planner prioritized_plan_warehouse.launch

roslaunch multi_robot_traj_planner prioritized_plan_random_env.launch

You can configure the simulation settings in the launch files.

(1) Environment: The simulation environment is selected by argument 'replay_map'. The build-in maps are located in /mapfile.

(2) Mission: The mission of the robots is selected by argument 'mission'. The build-in mission files are located in /missions.

(3) Priority Assignment: A prioritized trajectory optimization method is applied to improve the computation efficiency. If argument 'plan_random_group' is true, the priority of the robots is randomly assigned. Otherwise, a novel priority assignment method proposed in the paper is applied.

(4) Backward Movement: If argument 'backward_enable' is true, the robots are able to move backward. Otherwise, the robots can only have forward speed.

Our implementation is built on top oflibMultiRobotPlanning andswarm_simulator. We thank Wolfgang Hönig and Jungwon Park for their great work.