Fast-LIO2 + SC-A-LOAM + Ego-Planner

使用Fast-LIO2 的激光惯性里程计，加入 SC-A-LOAM 实现回环闭合和位姿图优化，最后使用 ego-planner 作为规划器

This project is based on FAST_LIO_SLAM、livox2pointcloud and ego-planner-swarm. It has been modified and greatly improved by emNavi Technology.

实验配置

• 环境：ROS1 Noetic
• 激光雷达：Mid360
• IMU：Mid360自带的IMU
• 驱动包：Livox SDK2

环境配置

# Ceres 2.1.0
sudo apt-get -y install liblapack-dev libsuitesparse-dev libcxsparse3 libgflags-dev libgoogle-glog-dev libgtest-dev
wget -O ceres-solver.zip https://github.com/ceres-solver/ceres-solver/archive/refs/tags/2.1.0.zip
unzip -q ceres-solver.zip -d "${TRDPARTY_DIR}"
pushd "${TRDPARTY_DIR}/ceres-solver-2.1.0"
mkdir build
cd build
cmake -DBUILD_SHARED_LIBS=TRUE ..
make -j8
sudo make install

# Eigen 3.3.7
wget -O eigen3.zip <https://gitlab.com/libeigen/eigen/-/archive/3.3.7/eigen-3.3.7.zip>
unzip -q eigen3.zip -d "${TRDPARTY_DIR}"
pushd "${TRDPARTY_DIR}/eigen-3.3.7"
mkdir build
cd build
cmake -DBUILD_SHARED_LIBS=TRUE ..
sudo make install
sudo ln -s /usr/include/eigen3/Eigen /usr/include/Eigen

# GTSAM
cd ~
git clone https://github.com/borglab/gtsam.git
mkdir build && cd build
cmake -D GTSAM_USE_SYSTEM_EIGEN=ON ..
make
sudo make install

# Livox SDK2
git clone https://github.com/Livox-SDK/Livox-SDK2.git
cd Livox-SDK2
cd build && cmake ..
make
sudo make install

# 下载 & 编译项目
git clone https://github.com/emNavi/Fast-LIO2.git

cd Fast-LIO2
catkin_make

使用

Mid360激光雷达上电，插入Mid360激光雷达网口至电脑，配置好 livox_ros_driver2 驱动包中的 IP 地址，确保雷达可使用

# 开启 Fast_LIO2 算法
bash ./run_fast_lio.sh 

# 开启 Ego-Planner 算法（需要去预设规划航点）
bash ./run_ego_planner.sh

注：预设规划航点可修改该文件