| 4,499 | 79 | 132 |
| 下载次数 | 被引频次 | 阅读次数 |
多传感器信息融合是实现无人驾驶的核心技术，多个传感器之间协同收集车辆周围环境的数据信息，经过多传感器融合结构的转换和处理，使用融合算法进行联合分析，能够使车辆全面地感知驾驶环境，帮助车辆完成自主导航、变道、控制速度等智能决策。基于多传感器信息融合的基本定义，从功能模型和结构模型介绍多传感器信息融合的基本形式；重点梳理多传感器信息融合的算法，分为随机类和人工智能类两个大类，详细分析各方法的原理及特点；最后总结出多传感器信息融合策略在实际应用时的主要步骤，同时分析其在无人驾驶场景中的应用，为多传感器信息融合未来理论研究方向和应用实践方向提供参考，从而完成多传感器信息融合的综合分析。
Abstract：Multi-sensor information fusion is the core technology to realize unmanned driving. Multiple sensors cooperate to collect the data information of the surrounding environment of the vehicle. After the transformation and processing of the multi-sensor fusion structure, the fusion algorithm is used for joint analysis, which enables the vehicle to perceive comprehensively the driving environment, and helps the vehicle complete intelligent decisions such as autonomous navigation, lane change, and control the speed. This paper first outlines the basic definition of multi-sensor information fusion, and introduces the basic forms of multi-sensor information fusion from functional models and structural models. Then, it focuses on sorting out the algorithms of multi-sensor information fusion,which are divided into two categories: random and artificial intelligence, the principle and characteristics of each method are detailed.Finally, the main steps of the multi-sensor information fusion strategy in practical application are summarized, and its application in the unmanned driving scene is analyzed at the same time, which provides a reference for the future theoretical research and application practice of multi-sensor information fusion. Reference, so as to complete the comprehensive analysis of multi-sensor information fusion.
[1]张庆，夏天，范轶飞，等.基于多传感器融合的运动目标跟踪算法[J].现代电子技术,2017,40(3):43-46.
[2]孙宁，秦洪懋，张利，等.基于多传感器信息融合的车辆目标识别方法[J].汽车工程,2017,39(11):1310-1315.
[3]于慧伶，梁玉亮，庞振男，等.基于多传感器融合的珍稀树木立体监测系统设计[J].东北林业大学学报,2016,44(12):91-93.
[4]李长勇，蔡骏，房爱青，等.多传感器融合的机器人导航算法研究[J].机械设计与制造,2017,(5):238-240.
[5]欧阳旭朗，何文忠，鹿玮，等.立体边防监视系统与多传感器信息融合技术研究[J].半导体光电,2018,39(2):298-304.
[6]Siljak D D. Decentralized control of complex systems[M]. New York:Academic,1991.
[7]WHITE F E. Data Fusion Lexicon[C]//Technical Panel for C3in Joint Directors of Laboratories. San Diego:Naval Ocean Systems Center, 1991:1-16.
[8]Varshncy P K. Multisensor data fusion[J]. E1ectronics&Communication Engineering Journal,1997,9(6):245-253.
[9]Klein L A. Sensor and data fusion concepts and application[M].北京:北京理工大学出版社,2004.
[10]李洋,赵鸣,徐梦瑶,等.多源信息融合技术研究综述[J].智能计算机与应用,2019,9(5):186-189.
[11]VAPNIK V N. The NATURE of Statistical Learning Theory Technometrics[J],1997,38(4):409-409.
[12]Xu L J, Chen Y Z, Cui P Y. Improvement of D-S evidential theory in multisensor data fusion system[C]. Huangzhou:Proceeding of the 5th World Congress on Intelligent Control and Automation, 2004.
[13]DENG J, CZARNECKI K. MLOD:A multi-view 3D object detection based on robust feature fusion method[C]//2019 IEEE Intelligent Transportation Systems Conference(ITSC), IEEE, 2019.
[14]晃锐,张科,李言俊.一种基于小波变换的图像融合算法[J].电子学报,2004,32(5):750-753.
[15]SZEGEDY C, TOSHEV A, ERHAN D. Deep Neural Net-works for Object Detection[C]. Advances in Neural Information Processing Systems, 2013:2553-2561.
[16]VIOLA P A, JONES M J. Robust Real-Time Face Detec-tion[J].International Journal of Computer Vision, 2004, 57(2):137-154.
[17]常启瑜.多传感器融合的车辆检测与跟踪系统研究[D].湖南:湖南大学,2019.
[18]David L Mills, Internet Time Synchronization:The Network Time Protocol[J]. IEEE trans. Communications, Oct. 1991, 39:1482-1493.
[19]汪文俊,龚育昌,朱建明.基于UDP的局域网内时钟同步协议[J].计算机应用与软件,2007(05):133-135.
[20]张中昀.基于雷达与视觉传感器信息融合的车辆检测方法研究[D].江苏:江苏大学,2018.
[21]高茹涵.基于多传感器融合的自主换道控制策略研究[D].吉林:吉林大学,2020.
[22]王一博.多传感器数据融合的道路行驶车辆识别方法研究[D].南京:南京航空航天大学,2020.
[23]张银.基于EKF智能车辆多传感器融合定位算法研究[D].长沙:长沙理工大学,2018.
[24]张浩然.基于RNN的自动导航车辆多传感器融合状态估计研究[D].成都:电子科技大学,2020.
[25]赵新宇.基于多传感器信息融合技术的车辆运行状况监测理论与方法研究[D].湖南:湖南大学,2005.
[26]Ahmed M. Sallam.基于雷达和传感器融合建模的车辆安全系统[D].兰州:兰州理工大学,2019.
基本信息:
DOI：
中图分类号:TP212
引用信息:
[1]施晓东,杨世坤.多传感器信息融合研究综述[J].通信与信息技术,2022,No.260(06):34-41.
基金信息: