基于多源信息融合技术的汽车自动驾驶控制系统设计  

Design of Automotive Autonomous Driving Control System Based on Multi-source Information Fusion Technology

作　　者：张泽东 ZHANG Zedong(Anhui Automobile Vocational and Technical College,Hefei 230000,China)

机构地区：[1]安徽汽车职业技术学院,安徽合肥230000

出　　处：《佳木斯大学学报(自然科学版)》2025年第2期57-60,共4页Journal of Jiamusi University:Natural Science Edition

摘　　要：为更好地实现汽车的自动驾驶控制,研究基于多源信息融合技术来对汽车的自动驾驶控制系统中的障碍物识别进行优化。通过采用多传感器结合的方法来实现路面点云数据计算,并基于该计算结果,结合图神经网络对相应的数据进行匹配。实验结果显示,障碍物在15m处,融合方法的测量误差值分别为0.94mm, 0.51mm和0.72mm,随着距离的增大,融合方法的优势性能更加明显。由此说明,研究设计的基于多源信息融合技术的汽车自动驾驶控制设计方法可以有效地识别路面障碍信息,保证驾驶安全性。In order to better achieve autonomous driving control of automobiles,research is conducted on optimizing obstacle recognition in the autonomous driving control system of automobiles based on multi-source information fusion technology.Firstly,the study adopts a combination of multiple sensors to achieve road surface point cloud data calculation.Based on this calculation result,match the corresponding data using a graph neural network.The experimental results show that at a distance of 15m,the measurement errors of the fusion method are 0.94mm,0.51mm,and 0.72mm,respectively.As the distance increases,the advantage performance of the fusion method becomes more obvious.This indicates that the research and design of an automotive autonomous driving control design method based on multi-source information fusion technology can effectively identify road obstacle information and ensure driving safety.

关 键 词：信息融合 激光雷达 点云 聚类 自动驾驶 

分 类 号：TP24[自动化与计算机技术—检测技术与自动化装置]