四足机器人 | GEM(elevation map) + Fast_lio(odometry) 环境部署记录
由于项目和科研需要，经常需要在不同的平台上（例如我的台式机和项目pc）部署 GEM(elevation map) + Fast_lio(odomtry) 的环境，因此记录下安装过程和过程中每次都会出现的和解决方案。
1. GEM && GEM Dependencies install
GEM是ZJU Robotics Lab开源的建elevation map程序
1）Grid Map INSTALL
Addr: https://github.com/anybotics/grid_map
INSTALL：
cd catkin_ws/src
git clone https://github.com/anybotics/grid_map.git
cd ../
catkin_make
问题：
In file included from /home/ysc/hedian_lib/grid_map_ws/src/grid_map-master/grid_map_filters/src/ThresholdFilter.cpp:9:0:
/home/ysc/hedian_lib/grid_map_ws/src/grid_map-master/grid_map_filters/include/grid_map_filters/ThresholdFilter.hpp:14:10: fatal error: filters/filter_base.hpp: 没有那个文件或目录
#include 解决：
将grid_map中的所有的filter_base.hpp换成filter_base.h。如图
2）Kindr
Addr : https://github.com/anybotics/kindr
git clone https://github.com/ANYbotics/kindr.git
mkdir build
cd build
cmake .. -DUSE_CMAKE=true
sudo make install
3) PCL
ROS已装背景下安装PCL和VTK
测试可用PCL版本：PCL-1.8
4) Eigen
由于fast lio需要Eigen 3.3.9，因此可以在已有Eigen的基础上再装一个3.3.9版本，使用时切换版本
Eigen 3.3.9版本下载
#git clone https://github.com/libeigen/eigen.git
wget https://gitlab.com/libeigen/eigen/-/archive/3.3.9/eigen-3.3.9.zip
unzip eigen-3.3.9.zip
Eigen3多版本切换
4) CUDA
测试过的GEM可用的CUDA版本
CUDA Version 11.0.207
CUDA Version 10.2.89
首先检查cuda版本
cat /usr/local/cuda/version.txt
CUDA Version 10.2.89
查看CUDA计算能力（GEM中CUDA配置需要）
cd /usr/local/cuda-11.0/samples/1_Utilities/deviceQuery
2 sudo make
3 sudo ./deviceQuery
查询结果：
因此，本机CUDA计算能力
arch=compute_86,code=sm_86
5) GEM
cd catkin_workspace/src
git clone https://github.com/ZJU-Robotics-Lab/GEM.git
git clone https://github.com/ZJU-Robotics-Lab/slam_msg.git
git clone https://github.com/ANYbotics/kindr_ros.git
cd ../
catkin_make
问题1：
[ 1%] Built target actionlib_generate_messages_eus
In file included from /usr/include/eigen3/Eigen/Core:42:0,
from /home/ysc/gem_test_0422/src/elevation_mapping_periodical/elevation_mapping/cuda/gpu_process.cu:15:
/usr/local/cuda-10.2/include/math_functions.hpp:54:2: warning: #warning "crt/math_functions.hpp is an internal header file and must not be used directly. Please use cuda_runtime_api.h or cuda_runtime.h instead." [-Wcpp]
#warning "crt/math_functions.hpp is an internal header file and must not be used directly. Please use cuda_runtime_api.h or cuda_runtime.h instead."
^~~~~~~
In file included from /usr/local/cuda-10.2/include/cuda_runtime.h:115:0,
from <command-line>:0:
/usr/local/cuda-10.2/include/crt/common_functions.h:74:24: error: token ""__CUDACC_VER__ is no longer supported. Use __CUDACC_VER_MAJOR__, __CUDACC_VER_MINOR__, and __CUDACC_VER_BUILD__ instead."" is not valid in preprocessor expressions
#define __CUDACC_VER__ "__CUDACC_VER__ is no longer supported. Use __CUDACC_VER_MAJOR__, __CUDACC_VER_MINOR__, and __CUDACC_VER_BUILD__ instead."
^
/usr/local/cuda-10.2/include/crt/common_functions.h:74:24: note: in definition of macro ‘__CUDACC_VER__’
#define __CUDACC_VER__ "__CUDACC_VER__ is no longer supported. Use __CUDACC_VER_MAJOR__, __CUDACC_VER_MINOR__, and __CUDACC_VER_BUILD__ instead."
^~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~
/usr/local/cuda-10.2/include/crt/common_functions.h:74:24: error: token ""__CUDACC_VER__ is no longer supported. Use __CUDACC_VER_MAJOR__, __CUDACC_VER_MINOR__, and __CUDACC_VER_BUILD__ instead."" is not valid in preprocessor expressions
#define __CUDACC_VER__ "__CUDACC_VER__ is no longer supported. Use __CUDACC_VER_MAJOR__, __CUDACC_VER_MINOR__, and __CUDACC_VER_BUILD__ instead."
^
/usr/local/cuda-10.2/include/crt/common_functions.h:74:24: note: in definition of macro ‘__CUDACC_VER__’
#define __CUDACC_VER__ "__CUDACC_VER__ is no longer supported. Use __CUDACC_VER_MAJOR__, __CUDACC_VER_MINOR__, and __CUDACC_VER_BUILD__ instead."
^~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~
[ 1%] Built target tf_generate_messages_lisp
/
问题1解决：
cd /usr/local/cuda-10.2/include/crt/common_functions.h
comment out line 74
#define __CUDACC_VER__ "__CUDACC_VER__ is no longer supported. Use __CUDACC_VER_MAJOR__, __CUDACC_VER_MINOR__, and __CUDACC_VER_BUILD__ instead.
问题2
问题2解决
这是CUDA Cmakelist中NVCC编译设置没设置对
CUDA用的是NVCC编译器，编译过程中不会在catkin_make窗口报错，但是会在GEM运行CUDA过程中报错
应该根据CUDA deviceQuery 查询得到的结果修改arch=compute_80,code=sm_80
src/elevation_mapping_periodical/elevation_mapping/cuda/CMakeLists.txt
project(gpu) # required cmake version
cmake_minimum_required(VERSION 2.8) # packages
find_package(CUDA REQUIRED)
find_package(Eigen3 REQUIRED)
find_package(kindr)
# set(CUDA_NVCC_FLAGS -O3;-G;-g)
SET(CUDA_NVCC_FLAGS ${CUDA_NVCC_FLAGS};-O3;-std=c++14 -gencode arch=compute_80,code=sm_80)
# SET(CUDA_NVCC_FLAGS ${CUDA_NVCC_FLAGS};-O3;-std=c++14 -gencode arch=compute_75,code=sm_75)
#include_directories(${EIGEN3_INCLUDE_DIRS})
file(GLOB_RECURSE CURRENT_HEADERS *.h *.hpp *.cuh)
file(GLOB CURRENT_SOURCES *.cpp *.cu)
source_group("Include" FILES ${CURRENT_HEADERS})
source_group("Source" FILES ${CURRENT_SOURCES})
cuda_add_library(gpu SHARED ${CURRENT_HEADERS} ${CURRENT_SOURCES})
问题3
这个是在一台性能较差的NX上跑的GEM，一跑起来就挂掉了，因为处理器性能太差了，处理不过来这么多点
问题4解决
把src/elevation_mapping_periodical/elevation_mapping/src/ElevationMapping.cpp中对image的处理部分注释掉（image对点云着色），会省掉很多算力
如图
注释掉就能跑起来了
问题5
code ~/src/elevation_mapping_periodical/elevation_mapping_demos/config/robots/fxz_robot.yaml
camera_params_yaml: "/home/nickle/yq_intrinsic.yaml" 要改路径，这里源码里设置的是绝对路径
2. FAST-Lio && FAST-Lio Dependencies install
1） Livox-SDK
Livox-SDK安装说明
https://github.com/Livox-SDK/Livox-SDK
git clone https://github.com/Livox-SDK/Livox-SDK.git
cd Livox-SDK
cd build && cmake ..
make
sudo make install
2）livox_ros_driver
https://github.com/Livox-SDK/livox_ros_driver
git clone https://github.com/Livox-SDK/livox_ros_driver.git ws_livox/src
cd ws_livox
catkin_make
3）FAST -LIO
cd ~/$A_ROS_DIR$/src
git clone https://github.com/hku-mars/FAST_LIO.git
cd FAST_LIO
git submodule update --init
cd ../..
catkin_make
source devel/setup.bash
问题1
In file included from /home/ysc/GEM_ws/src/fast-lio/FAST_LIO_ws/src/FAST_LIO/src/IMU_Processing.hpp:11:0,
from /home/ysc/GEM_ws/src/fast-lio/FAST_LIO_ws/src/FAST_LIO/src/laserMapping.cpp:46:
/home/ysc/GEM_ws/src/fast-lio/FAST_LIO_ws/src/FAST_LIO/include/common_lib.h:8:10: fatal error: fast_lio/Pose6D.h: 没有那个文件或目录
#include 解决：
手动将livox_ros_driver生成的msg复制到ws/devel/include下
问题2
FASTlio适配的eigen是3.3.9
如果不是的话会产生大量的Eigen有关的报错，这时候可以按照上面的说明切换或者安装Eigen3-3-9