BUPT-ROBAN-SIM

北邮智能机器人交互实验仿真作业，仅供参考

Build ROBAN V-Rep SIM Environment on your PC

due to the v-rep sim is extremely slow in cpu, you can try to build the sim on you own pc. I have built the sim on my ubuntu 20.04 PC .

WSL (NOT TESTED)

TODO......

Ubuntu 20.04

TODO......

docker

程序运行使用说明（环境配置与参数设置等）

仿真环境

打开roscore后，启动V-Rep，打开ai_innovative_roban_sim_task10_upstair.ttt。

之后启动bodyhub和ikmodule：

sleep 3s
rosrun  ik_module ik_module_node &
. /home/fan/robot_ros_application/catkin_ws/devel/setup.bash
echo "121" | sudo -S bash bodyhub.sh &

启动定位和建图功能

分别在三个终端中启动：

# 终端1，转换图像格式
# 该脚本会将V-Rep中的图像转换为ros中可以使用的格式，以及转换深度图
cd ~/robot_ros_application/catkin_ws/src/ros_actions_node/scripts/game/2022/normal_sim_game/ai_innovative_roban_sim/scripts && python3 ./sim_image_convert_slam_form.py
# 终端2，转换坐标系
cd ~/fan/robot_ros_application/catkin_ws/src/ros_actions_node/scripts/game/2022/normal_sim_game/ai_innovative_roban_sim/scripts && python3 ./sim_pr_convert_pose.py 
# 终端3，打开octomap建图
roslaunch ros_actions_node sim_octomap.launch
# 终端4，通过rviz查看建图结果
rosrun rviz rviz -d roban.rviz

导航模块使用

进入 nav_dev，使用catkin_make指令编译。

cd nav_dev
catkin_make 
# or 'catkin build'

之后终端中输入

source devel/setup.bash
roslaunch humanoid_planner_2d humanoid_planner_2d.launch 

启动轨迹跟踪

cd  ~/robot_ros_application/catkin_ws/src/ros_actions_node/scripts/game/2022/caai_roban_challenge/path_track && python3 Task_path_tracking.py 

启动键盘控制

cd  ~/robot_ros_application/catkin_ws/src/ros_actions_node/scripts/game/2022/caai_roban_challenge/path_track && python3 key_ctrl.py 

Known Issues

1. The pose takes the truth-value from sim. You need to run a slam system like ORB-SLAM or VINS-Fusion to get odom when you deploy the code to the real robot.
2. This project only use the path plan, not considering the footstep planning. You can try the footstep planning from the humanoid_navigation to achieve a better motion.

Reference

humanoid_navigation : ROS stack with footstep planning and localization for humanoid robots