基于深度学习的道路自动化监测系统及其应用研究
Road Automation Monitoring System Based on Deep Learning:an Application Research
柳士伟 1王荣 2张同虎 3吴回国4
作者信息
- 1. 空军后勤部工程代建管理办公室第一代建项目部,江苏南京 210018
- 2. 空后代建办第一代建部计划发包室,江苏南京 210000
- 3. 灌南县水利局,江苏连云港 222500
- 4. 河海大学,江苏南京 210024
- 折叠
摘要
在道路施工监测中,传统的监测方法效率低,无法实时连续准确地预测土体变形.提出一种集成人工智能技术的道路自动化监测系统,该系统由实时物联网系统和数据处理系统组成.实时物联网系统包括双压力传感器埋入式沉降仪、数据采集系统和网络传输系统;数据处理系统则利用深度学习算法对实测数据进行训练,实现土体变形的预测.介绍了该监测系统的构成和工作原理,通过现场试验对该监测系统进行验证,将双压力传感器埋入式沉降仪的数据与沉降板的数据进行对比分析,结果显示两者之间的误差仅为 6.7%,表明自动化监测仪器在道路施工监测中具有高精度.同时,现场试验结果还证明了基于深度学习算法的变形预测方法能够准确地对道路施工过程中的土体变形进行预测,其预测最大误差仅为 5.3%.
Abstract
Traditional monitoring methods are inefficient and cannot predict soil deformation in real-time and continuously with accuracy in road construction.A road automation monitoring system integrated with artificial intelligence technology was pro-posed.The system consists of a real-time Internet of Things(IoT)system and a data processing system.The real-time IoT system in-cludes embedded settlement instruments with dual pressure sensors,a data acquisition system,and a network transmission system.The data processing system utilizes deep learning algorithms to train the measured data to predict soil deformation.The composition and working principles of the monitoring system was introduced.The system was validated through on-site experiments.By comparing the data from the dual pressure sensors in the embedded settlement instruments with the data from settlement plates,the results show that the error between the two is only 6.7%.This indicated that the automated monitoring instrument has high precision in road construc-tion monitoring.On-site experimental results also prove that the deformation prediction method based on deep learning algorithms can accurately predict soil deformation during the road construction process,with a maximum prediction error of only 5.3%.
关键词
道路自动化监测/人工智能/物联网/深度学习/土体变形Key words
road automation monitoring/artificial intelligence/Internet of Things(IoT)/deep learning/soil deformation引用本文复制引用
出版年
2024
NETL
NSTL
万方数据