基于多传感器融合的电力管廊3D建图研究  

Power pipe corridor 3D mapping research based on multi-sensor fusion

作　　者：李乾 谢正宽 曹玉超 王明明 孙晓云 LI Qian;XIE Zhengkuan;CAO Yuchao;WANG Mingming;SUN Xiaoyun(State Grid Corporation of China Hebei Province Shijiazhuang Power Supply Branch,Shijiazhuang 050000,China;School of Electrical and Electronic Engineering,Shijiazhuang Tiedao University,Shijiazhuang 050043,China)

机构地区：[1]国网河北省有限公司石家庄供电分公司,河北石家庄050000 [2]石家庄铁道大学电气与电子工程学院,河北石家庄050043

出　　处：《传感器与微系统》2025年第1期21-24,共4页Transducer and Microsystem Technologies

基　　金：国网河北省电力有限公司科技项目(5GHE5M00DL52250629)。

摘　　要：针对城市地下管廊、电缆隧道的复杂工况环境下,利用单一传感器进行环境建图不能准确构建复杂环境地图的问题,提出一种改进RTAB-MAP算法融合多种传感器进行环境建图的方法。实验在装有机器人操作系统(ROS)的智能小车上进行,采用激光雷达(LiDAR)和RGB-D深度相机融合建图。实验结果表明:相比于传统的RTAB-MAP算法,融合之后的方法在建图过程中建图精度比仅使用激光雷达建图小了0.114 m,比仅使用RGB-D相机建图小了0.128 m。Aiming at the problem that a single sensor can not accurately construct complex environment map in complex working conditions of urban underground pipe corridor and cable tunnel,an improved real-time appearance-based mapping(RTAB-MAP)algorithm fusing multiple sensors for environment mapping is proposed.The experiment is carried out on an intelligent car equipped with robot operating sysytem(ROS)system,using fusion of LiDAR and RGB-D depth camera for mapping.The experimental results show that compared with the traditional RTAB-MAP algorithm,the mapping precision of the fusion method is 0.114 m smaller than that only using LIDAR and 0.128 m smaller than that using only RGB-D camera in the process of mapping.

关 键 词：城市地下管廊 RTAB-MAP算法 同步定位和地图构建 多传感器融合 机器人操作系统 

分 类 号：TP242.6[自动化与计算机技术—检测技术与自动化装置]