先进的自动驾驶数据采集与存储技术研究综述  

Review on Advanced Autonomous Driving Data Acquisition and Storage Technology

作　　者：赵浩源 刘金来 禹梓浩 武涛[1] 孙天骏 Zhao Haoyuan;Liu Jinlai;Yu Zihao;Wu Tao;Sun Tianjun(Jilin University,Changchun 130022;State Key Laboratory of Automobile Simulation and Control,Jilin University,Changchun 130022)

机构地区：[1]吉林大学,长春130022 [2]吉林大学,汽车仿真与控制国家重点实验室,长春130022

出　　处：《汽车技术》2024年第2期1-16,共16页Automobile Technology

基　　金：国家级大学生创新创业训练项目(202310183133);汽车动力传动与电子控制湖北省重点实验室开放基金项目(ZDK1202305)。

摘　　要：针对自动驾驶数据采集过程中的触发控制与存储记录问题,从先进的自动驾驶数据采集与存储技术出发,首先调研了国内外部分整车制造商车辆的传感器配置方案,然后对汽车事件数据记录(EDR)系统、自动驾驶数据记录系统(DSSAD)以及一般L3级自动驾驶汽车的数据采集与存储方案展开综述,得出了现有数据采集与存储技术将会产生低质量、高噪声的冗余数据的结论,最后基于特斯拉影子模式介绍了一种人机共驾模式下的驾驶数据采集存储方式,总结归纳出目前先进的自动驾驶数据采集技术研究的发展趋势。In view of the trigger control and storage-recording in the process of automatic driving data collection,this paper started from advanced automatic driving data acquisition technology and advanced automatic driving storage technology,firstly investigated the sensor configuration methods of China and foreign automakers,then reviewed the automobile Event Data Recorder(EDR)system,autonomous driving data recording system and general L3 level autonomous vehicle data acquisition and storage schemes,and concluded that the existing data acquisition and storage technology could generate low quality and high noise redundant data.Finally,based on Tesla’s Shadow Mode,the paper introduced a driving data acquisition and storage method for man-machine co-driving mode,and summarized the development trend of advanced autonomous driving data acquisition technology.

关 键 词：自动驾驶 数据采集 数据存储 影子模式 

分 类 号：TP391.4[自动化与计算机技术—计算机应用技术]