安装ubuntu18.04：
本机器型号安装ubuntu没有什么疑难杂症,跟着网上大佬的教程，下载镜像文件，做启动盘安装即可。
分区方案：
efi 2G
swap 16G
/ 500G
启动引导选择efi的那个分区
有独显的记得进入编辑启动项 在quite splash 后加 =nomodeset
按F10 进入 try ubuntu 然后安装。
换源
sudo gedit /etc/apt/sources.list
把原文档用下列内容替换
deb https://mirrors.tuna./ubuntu/ bionic main restricted universe multiversedeb-src https://mirrors.tuna./ubuntu/ bionic main restricted universe multiversedeb https://mirrors.tuna./ubuntu/ bionic-updates main restricted universe multiversedeb-src https://mirrors.tuna./ubuntu/ bionic-updates main restricted universe multiversedeb https://mirrors.tuna./ubuntu/ bionic-backports main restricted universe multiversedeb-src https://mirrors.tuna./ubuntu/ bionic-backports main restricted universe multiversedeb https://mirrors.tuna./ubuntu/ bionic-security main restricted universe multiversedeb-src https://mirrors.tuna./ubuntu/ bionic-security main restricted universe multiversedeb https://mirrors.tuna./ubuntu/ bionic-proposed main restricted universe multiversedeb-src https://mirrors.tuna./ubuntu/ bionic-proposed main restricted universe multiverse
禁用nouveau
sudo gedit /etc/modprobe.d/blacklist.conf
在最后一行加入
blacklist nouveauoptions nouveau modeset=0
更新，重启
sudo update-initramfs -ureboot
装N卡驱动
sudo add-apt-repository ppa:graphics-drivers/ppasudo apt updateubuntu-drivers devices
查看本机可装显卡驱动，推荐安装最后有recommend的那个。
sudo apt install nvidia-driver-440
如果该命令无法安装，添加源后再次执行
deb /ubuntu/ trusty main universe restricted multiverse
装完检查
df -h
如果几个Loop都是100%
执行
sudo apt autoremove --purge snapd
清除后再次检查。
重启，一般来说能直接进ubuntu就没有问题了
检查驱动
nvidia-smi
装pip
sudo apt updatesudo apt-get install python-pipsudo apt-get install python3-pip
装完之后检查
pip --versionpip3 --version
pip换源
mkdir ~/.pipcd ~/.pip && touch pip.confsudo gedit pip.conf
复制粘贴下面
[global]index-url = /simple[install]use-mirrors =truemirrors =/simple/trusted-host =
装chrome
百度搜索chrome，官网下载
CUDA+cudnn+装pytorch
一定要先去pytorch官网看，能直接用pip下载的需要CUDA几！！
自己下包实在是太慢了！
gcc g++降级（有的可以不用）
sudo apt-get install gcc-4.8sudo apt-get install g++-4.8ls -l gcc* sudo mv gcc gcc.bak #备份sudo ln -s gcc-4.8 gcc #重新链接ls -l g++*sudo mv g++ g++.baksudo ln -s g++-4.8 g++gcc -v g++ -v #查看版本号
装CUDA
官网下载
sudo sh cuda_10.1.130_410.48_linux.run（文件名自己换） --no-opengl-libs
双显卡一定要加–no-opengl-libs 不然会在登陆界面无限循环
10.0以前的版本是加–no-opengl-files
安装完成后 在bashrc文件内加上
export CUDA_HOME=/usr/local/cuda（或者cuda10.1 看自己版本） export PATH=$PATH:$CUDA_HOME/bin export LD_LIBRARY_PATH=/usr/local/cuda-10.0/lib64${LD_LIBRARY_PATH:+:${LD_LIBRARY_PATH}}
测试
测试：cd /usr/local/cuda/samples/1_Utilities/deviceQuery sudo make./deviceQuery
最后result=pass 成功。
安装CUDNN
官网注册下载
tar -zxvf cudnn-10.1-linux-x64-v7.6.5.32（文件名自己改）.tgzsudo cp cuda/include/cudnn.h /usr/local/cuda/includesudo cp cuda/lib64/libcudnn* /usr/local/cuda/lib64sudo chmod a+r /usr/local/cuda/include/cudnn.h sudo chmod a+r /usr/local/cuda/lib64/libcudnn*sudo dpkg -i libcudnn7_7.6.5.32-1+cuda10.1_amd64.debsudo dpkg -i libcudnn7-dev_7.6.5.32-1+cuda10.1_amd64.debsudo dpkg -i libcudnn7-doc_7.6.5.32-1+cuda10.1_amd64.deb
测试：
cp -r /usr/src/cudnn_samples_v7 $HOMEcd $HOME/cudnn_samples_v7/mnistCUDNNmake clean && make./mnistCUDNN
test passed 成功。
安装tensorflow
装tensorflowpip3 install tensorflow-gpu==1.13.1
测试：
python3 import tensorflow as tf hello=tf.constant("Hello,Tensorflow!") sess=tf.Session() print(sess.run(hello))
安装ros
设置源
sudo sh -c '. /etc/lsb-release && echo "deb http://mirrors.tuna./ros/ubuntu/ $DISTRIB_CODENAME main" > /etc/apt/sources.list.d/ros-latest.list'
加key
sudo apt-key adv --keyserver 'hkp://:80' --recv-key C1CF6E31E6BADE8868B172B4F42ED6FBAB17C654
更新
sudo apt-get update
安装ros和一些工具包
sudo apt install ros-melodic-desktop-fullsudo apt-get install ros-melodic-rqtsudo apt-get install ros-melodic-rqt-common-plugins
装完后在bashrc文件添加
source /opt/ros/melodic/setup.bash
建立工作空间
在~目录下
mkdir catkin_wscd catkin_wsmkdir srccatkin_make
安装rbx1
sudo apt-get install ros-melodic-arbotix ros-melodic-openni-camera \ros-melodic-rosbridge-suite \ros-melodic-rgbd-launch \ros-melodic-turtlebot-* ros-melodic-kobuki-* ros-melodic-moveit-python \python-pygraph python-pygraphviz python-easygui \mini-httpd ros-melodic-laser-pipeline ros-melodic-ar-track-alvar \ros-melodic-laser-filters \ros-melodic-depthimage-to-laserscan ros-melodic-shape-msgs \ros-melodic-gazebo-ros ros-melodic-gazebo-ros-pkgs \ros-melodic-gazebo-msgs ros-melodic-gazebo-plugins \ros-melodic-gazebo-ros-control ros-melodic-cmake-modules \ros-melodic-smach ros-melodic-smach-ros ros-melodic-grasping-msgs \ros-melodic-executive-smach ros-melodic-smach-viewer \ros-melodic-tf2-web-republisher \ros-melodic-move-base-msgs ros-melodic-fake-localization \graphviz-dev libgraphviz-dev gv python-scipy liburdfdom-tools \ros-melodic-ar-track-alvar* \ros-melodic-map-server ros-melodic-move-base* \ros-melodic-simple-grasping
cd catkin_wscd srcgit clone /pirobot/rbx1.git或者保存到自己的gitee下gitgit clone /different_way/rbx1.gitcd ..catkin_make
装kinect1驱动以及ros接口
在~目录下
git clone /different_way/kinect_driver.git
跟着里面的readme安装
还是在~目录下
git clone /different_way/libfreenect.gitcd libfreenectmkdir buildcd buildcmake ..makesudo make install
在工作空间的src文件夹下
git clone /different_way/freenect_stack.gitcd ..catkin_make
双系统时间不对解决办法
在ubuntu下
安装时间校准服务
sudo apt-get install ntpdatesudo ntpdate sudo hwclock --localtime --systohc
