UWB数据集分析

UWB

超宽带（Ultra Wide Band，UWB）技术是一种无线载波通信技术，它不采用正弦载波，而是利用纳秒级的非正弦波窄脉冲传输数据，因此其所占的频谱范围很宽。
UWB技术具有系统复杂度低，发射信号功率谱密度低，对信道衰落不敏感，截获能力低，定位精度高等优点，尤其适用于室内等密集多径场所的高速无线接入。

NoopLoop数据分析

使用NoopLoop公司的UWB节点，在塔克小车上采集数据，需要将数据集中的测距信息导出。本次实验使用Faster-Lio作为小车定位。

• 录制rosbag
  rosbag中使用rosbag info查看数据集内容：

  path:        uwb
  version:     2.0
  duration:    36.9s
  start:       Mar 30 2024 17:40:30.91 (1711791630.91)
  end:         Mar 30 2024 17:41:07.78 (1711791667.78)
  size:        254.1 MB
  messages:    18347
  compression: none [275/275 chunks]
  types:       dynamic_reconfigure/Config            [958f16a05573709014982821e6822580]
              dynamic_reconfigure/ConfigDescription [757ce9d44ba8ddd801bb30bc456f946f]
              geometry_msgs/Twist                   [9f195f881246fdfa2798d1d3eebca84a]
              livox_ros_driver2/CustomMsg           [e4d6829bdfe657cb6c21a746c86b21a6]
              nav_msgs/Odometry                     [cd5e73d190d741a2f92e81eda573aca7]
              nlink_parser/LinktrackNodeframe2      [5e015baafc881617fa57f6ba7fd27747]
              rosgraph_msgs/Log                     [acffd30cd6b6de30f120938c17c593fb]
              sensor_msgs/Imu                       [6a62c6daae103f4ff57a132d6f95cec2]
              sensor_msgs/PointCloud2               [1158d486dd51d683ce2f1be655c3c181]
              std_msgs/Float32                      [73fcbf46b49191e672908e50842a83d4]
              tf2_msgs/TFMessage                    [94810edda583a504dfda3829e70d7eec]
  topics:      /Odometry                                366 msgs    : nav_msgs/Odometry                    
              /bat_vol                                1831 msgs    : std_msgs/Float32                     
              /cloud_registered                        366 msgs    : sensor_msgs/PointCloud2              
              /cloud_registered_body                   366 msgs    : sensor_msgs/PointCloud2              
              /cmd_vel                                   2 msgs    : geometry_msgs/Twist                  
              /imu                                    1832 msgs    : sensor_msgs/Imu                      
              /livox/imu                              7334 msgs    : sensor_msgs/Imu                      
              /livox/lidar                             367 msgs    : livox_ros_driver2/CustomMsg          
              /nlink_linktrack_nodeframe2             1834 msgs    : nlink_parser/LinktrackNodeframe2     
              /odom                                   1831 msgs    : nav_msgs/Odometry                    
              /rosout                                   17 msgs    : rosgraph_msgs/Log                     (2 connections)
              /rosout_agg                                2 msgs    : rosgraph_msgs/Log                    
              /tarkbot_robot/parameter_descriptions      1 msg     : dynamic_reconfigure/ConfigDescription
              /tarkbot_robot/parameter_updates           1 msg     : dynamic_reconfigure/Config           
              /tf                                     2197 msgs    : tf2_msgs/TFMessage                    (2 connections)

  其中，/Odometry是定位信息，/nlink_parser/LinktrackNodeframe2是测距信息。
• 编译msg文件

  mkdir -p catkin_uwb/src
  cd catkin_uwb/src
  git clone --recursive https://github.com/nooploop-dev/nlink_parser.git
  cd ..
  catkin_make

• 使用Python和rosbag play实现数据回放和接收

  import rospy
  from nav_msgs.msg import Odometry
  from nlink_parser.msg import LinktrackNodeframe2
  
  rospy.init_node('uwb_test')
  
  data = []
  data_single = []
  init_time = 0.0
  init = False
  
  def odom_callback(msg: Odometry):
      global data_single, init, init_time
      if not init:
          init_time = msg.header.stamp.to_sec()
      init = True
      # print(type(msg.header.stamp.to_sec()))
      data_tmp = [msg.header.stamp.to_sec() - init_time,
                  msg.pose.pose.position.x,
                  msg.pose.pose.position.y,
                  msg.pose.pose.position.z]
      data.append([data_tmp, data_single])
  
  def uwb_callback(msgs):
      global data_single
      # print(len(msg.nodes))
      # print(msg.nodes[0])
      # print(msg.nodes[1])
      data_single = []
      for msgs in msgs.nodes:
          data_single.append(msgs.dis)
  
  if __name__ == '__main__':
      rospy.Subscriber('/Odometry', Odometry, odom_callback)
      rospy.Subscriber('/nlink_linktrack_nodeframe2', LinktrackNodeframe2, uwb_callback)
      # rospy.spin()
      while not rospy.is_shutdown():
          # print(data)
          pass
      with open('result.csv', 'w') as f:
          f.writelines('time,x,y,z,uwb1,uwb2,uwb3,uwb4\n')
          for da in data:
              f.writelines(str(da[0][0]) + ',' +
                          str(da[0][1]) + ',' +
                          str(da[0][2]) + ',' +
                          str(da[0][3]) + ',')
              for sin in da[1]:
                  f.writelines(str(sin) + ',')
              f.writelines('\n')

  在程序目录下会出现result.csv