摘 要:为了解决高速公路隧道内感知设备的相互独立,安全管理靠监控员及养护人员进行人工巡检等效率低下问题,将多元感知设备通过反馈控制的闭环模型进行数据融合,应用 Dempster-Shafer 证据理论将分立数据及融合数据进行预测和判断,通过模型训练得出最佳数据融合模型。开发出基于多元感知及多源数据融合的隧道安全监测与预警系统,并应用于紫惠高速的好义隧道的安全管理,大大提高管理效率。
关键词:多元感知;数据整合;隧道安全;Dempster-Shafer 证据理论;卷积神经网络
DOI:10.19850/j.cnki.2096-4706.2022.19.034
中图分类号:TP212;U491.2 文献标识码:A 文章编号:2096-4706(2022)19-0142-04
Research on Tunnel Security Monitoring and Early Warning System Based on Multivariate Perception and Data Fusion Technology
LI Yuhuan, LI Yujiang, HU Cuiyun
(Guangdong Feida Traffic Engineering Co., Ltd., Guangzhou 510663, China)
Abstract: In order to solve the inefficient problems that the mutual independence of the perception equipment in the highway tunnel, the safety management relies on the monitor and maintenance personnel to manually inspect and so on, the multiple perception equipment is integrated data through the closed-loop model of feedback control, and the Demoster-Shafer evidence theory is applied to predict and judge the discrete data and fusion data, and the best data fusion model is obtained through model training. A tunnel security monitoring and early warning system based on multiple perception and multi-source data fusion is developed, and applied to the security management of The Haoyi Tunnel of Zihui Expressway, which greatly improves the management efficiency.
Keywords: multivariate perception; data integration; tunnel security; Dempster-Shafer evidence theory; convolutional neural network
参考文献:
[1] 罗俊海,王章静 . 多源数据融合和传感器管理 [M]. 北京:清华大学出版社,2015.
[2] 高春艳,卢建,张明路,等 . 基于多传感器数据融合的多目标跟踪算法研究 [J]. 制造业自动化,2022,44(7):95-97+101.
[3] 王玉婷,李静,蔡晓禹 . 基于改进 D-S 证据理论数据融合的路段单元交通状态判别方法 [J]. 交通运输研究,2021,7(6):31-39.
[4] 丁春山 . 传感器管理技术研究现状与展望 [J]. 系统工程与电子技术,2020,42(12):2761-2770.
[5] 谢婷婷 . 基于卷积神经网络的两阶段图像分类研究 [D].成都:电子科技大学,2022.
作者简介:李玉环(1974.11—),女,汉族,江西宜春人,高级工程师,硕士研究生,研究方向:智慧交通。