iphet / ego-planner-for-ground-robot

[toc]

Environment Configuration

qpOASES

We use qpOASES to solve MPC，and MPC is used for trajectory tracking

qpOASES solver installation :

cd qpOASES
mkdir build
 cd build
 cmake ..
 sudo make
 sudo make install

patchwork

patchwork is used for ground segmentation. This planner accepts 3d point cloud information Pointcloud2. After the point cloud is segmented on the ground, the point cloud is projected to the 2d plane to construct a 2d grid map for planning.

Follow the official repository to download this package：patchworl_github

Simulation environment

The simulation environment uses the open source simulation environment of the CMU Robotics Institute. Users need to install the relevant libraries and download the mesh file. We recommend to use the indoor environment to test the planner.

The environment have two versions for ros-melodic and ros-noetic. The package in this project is for ros-melodic.

CMU-environment

Quick Start

First , start the simulation environment :

roslaunch vehicle_simulator system_indoor.launch

Second, start the planner :

roslaunch ego_planner run_in_sim.launch

In rviz , use 2d nav goal to set the goal :


If you can't see the imag in this file, you can see it in the folder "imag".

THANKS

We are very grateful for Fei Gao and the EGO-Planner he proposed. Our algorithm is implemented under their algorithm framework. If you are interested in our algorithm, you may see the origin algorithm in :

EGO-Planner