Skip to content

PolarisXQ/SCURM_SentryNavigation

Folders and files

NameName
Last commit message
Last commit date

Latest commit

 

History

15 Commits
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 

Repository files navigation

SCURM火锅战队 24~25赛季哨兵导航

开源的初衷是想把自己的一些好的想法和大家一起分享,所以欢迎各位联系我和我讨论,欢迎PR,提issue🫠~

主要的创新:

  1. 实现FAST-LIO2的重定位模式,不需要另外运行重定位算法。算力需求小,使得整个框架在NUC12上的资源占用在30%左右;精度高,定位误差小;内存占用不会随着时间增长。

  2. 改进navigation2的故障恢复行为,卡住时机器人会向无碰撞的方向运动。

算法框架和思路详见技术报告

Gazebo仿真指路👉SCURM_Nav_Tutorial

Docker镜像使用方法参阅DevcontainterGuide

包说明

Package Name Description
✅auto_aim_interfaces 自瞄接口
autonomous_exploration_development_environment 地形分析包terrain_analysis和terrain_analysis_ext,其他的是小工具无关紧要
BehaviourTree.CPP MODIFIED BehaviourTree lib
✅cmd_chassis twist2chassis_cmd:将twist加上底盘的控制方式(如是否小陀螺),发出到串口接收的话题;
twist_transformer, fake_joint用于实现底盘到云台的速度解算
✅control_panel 模仿裁判系统发消息
✅FAST_LIO 修改版fastlio,具备建图和重定位功能(须配合icp_relocalizatiion使用)
✅icp_relocalization 基于icp实现的重定位,须配合修改版FAST_LIO使用
✅livox_ros_driver2 livox雷达驱动
✅nav2_plugins
- behavior_ext_plugins
- costmap_intensity
Costume nav2 plugins
- an enhenced back_up action that move toward free space
- 2 costmap_2d layer that use intensity filed of pointcloud msg rather than height (use with terrain analysis in autonomous_exploration_development_environment)
✅rm_decision_cpp 烧饼决策系统
✅rm_interfaces 通讯协议
✅sentry_bringup 哨兵启动文件
✅sentry_description 烧饼urdf

LAUNCH

MAPPING

  • launch mapping node
ros2 launch sentry_bringup mapping.launch.py
  • save map
# occupancy grid map
ros2 run nav2_map_server map_saver_cli -t /projected_map -f test_map --fmt png
# save pcd
ros2 service call /map_save std_srvs/srv/Trigger
  • then terminate all nodes, pcd file will be saved in /PCD/scans.pcd

MAP PROCESSING(使用CloudCompare删除建图中的人物残影, 如不处理可以跳过,但记得按照最后一步的提示替换掉pcd文件)

process pcd file

  • drag the pcd file to CloudCompare(globalmap or scans.pcd), select the pointcloud in the left panel, then tools->clean->SOR filter, set the parameters (25,1 is a baseline) and apply
  • select the processed pointcloud from last step, then tools->segmentation->Label Connected Components, set the parameters and apply
  • pick out the CC#0(ususally this one), then tools->Other->Remove duplicate points, keep 1 point per 0.01-0.1m to reduce the size of the pointcloud
  • select the processed pointcloud, then file->save as, select .pcd format
  • replace the original pcd file with the processed one. Globally seach and replace '/home/sentry_ws/src/sentry_bringup/maps/GlobalMap.pcd' with '/path/to/your/map'

LAUNCH ALL

ros2 launch sentry_bringup bringup_all_in_one.launch.py

About

No description, website, or topics provided.

Resources

Stars

Watchers

Forks

Releases

No releases published

Packages

No packages published

Contributors 2

  •  
  •