轨道车辆检测系统点云数据仿真方法研究  

Research on Point Cloud Data Simulation Method of Rail Vehicle Detection System

作　　者：张建豪 梁晨锐 ZHANG Jianhao;LIANG Chenrui(State Key Laboratory of Traction Power,Southwest Jiaotong University,Chengdu 610031,China)

机构地区：[1]西南交通大学牵引动力学国家重点实验室,四川成都610031

出　　处：《机械制造与自动化》2024年第6期133-137,共5页Machine Building & Automation

摘　　要：激光雷达可以通过激光脉冲对目标区域进行密集采样来快速生成三维点云,三维点云数据能够有效地用于环境三维重建、障碍物检测等算法,但轨道车辆在地铁环境下,采集到符合算法要求的激光雷达数据是一项非常具有挑战性的任务,需要消耗大量人力、物力。针对以上情况,提出一种轨道车辆检测系统点云数据仿真方法,通过对地铁隧道和激光雷达传感器建模,利用激光雷达工作原理生成有向激光束,并将有向激光束通过雷达模型和隧道模型的位置关系转换到相同坐标系中,通过光线投射原理来获得仿真数据。基于该方法,对不同半径线路以及含有障碍物的工况进行数据仿真。试验结果表明:该数据仿真方法有助于轨道车辆检测系统应用激光雷达的算法开发。Although LIDAR can quickly generate 3D point clouds by densely sampling the target area with laser pulses,and 3D point cloud data can be effectively used for environmental reconstruction,obstacle detection and other algorithms,LIDAR data collecting in line with the algorithm requirements for rail vehicles in subway environments is a huge challenge requiring enormous human and material resources.To address the issue,this thesis proposes a point cloud data simulation method for rail vehicle detection systems.By modeling the subway tunnel and LIDAR sensor,directed laser beams are generated using the working principle of LIDAR,and the directed laser beams are transformed into the same coordinate system through the position relationship between the LIDAR model and the tunnel model.Simulation data is obtained through the principle of ray projection.Based on this method,data simulation for different radius lines and working conditions with obstacles are conducted,and the experimental results show that the proposed data simulation method will help the algorithm development of rail vehicle detection systems using LIDAR.

关 键 词：激光雷达 三维点云 隧道模型 激光雷达模型 数据仿真 

分 类 号：TP391.9[自动化与计算机技术—计算机应用技术]