基于AI视觉和激光雷达数据融合技术的主动障碍物检测系统研究
Research on active obstacle detection system based on AI Vision and LiDAR data fusion technology
陆源清1
作者信息
- 1. 无锡地铁运营有限公司,江苏无锡 214000
- 折叠
摘要
随着地铁运行效率提升需求的日益增长,越来越多的地铁运营线路开始采用FAO运营模式.FAO的引入使得传统的被动式障碍物检测方式已无法满足行业发展需求,因此,列车主动障碍物检测系统应运而生.文章基于当前行业内某主流列车主动障碍物检测系统,通过深入剖析其系统架构、网络拓扑、系统功能及关键技术指标,重点研究该系统所依托的AI视觉和激光雷达数据融合技术、毫米波雷达技术和限界判断及检测技术相关工作原理,以及在实际应用过程中面临的应用技术难点,从而进一步探讨该技术在无锡地铁及其他更多城市地铁系统中的广阔应用前景.
Abstract
With the increasing demand for improving the efficiency of metro operations,more and more metro lines are adopting the fully automated operation(FAO)mode.The introduction of the FAO system has made traditional passive obstacle detection methods unable to meet the development needs of the industry.Therefore,active obstacle detection system for trains have emerged.This article focuses on a mainstream train active obstacle detection system in the contemporary industry.Through in-depth analysis of its system architecture,network topology,system functions,and key technical indicators,the article focuses on the research of the relevant working principles such as artificial intelligence(AI)vision and laser radar data fusion technology,millimeter wave radar technology,and clearance judgment and detection technology that the system relies on,and the application technical difficulties confronted in the actual application,thereby further explores the broad application prospect of this technology in Wuxi metro and other urban metro systems.
关键词
地铁/FAO/AI/主动障碍物检测系统/数据融合Key words
metro/FAO/AI/active obstacle detection system/data fusion引用本文复制引用
出版年
2024
NETL
NSTL
万方数据