基于高速通信的港口设备远程检测与控制技术研究  

Design of remote detection and control technology for port equipment based on high-speed mobile communication

作　　者：徐晓强 丁峰 毕淑敏 XU Xiaoqiang;DING Feng;BI Shumin(Wuhu Port and Businss Co.,Ltd.,Anhui Wuhu 241000,China)

机构地区：[1]芜湖港务有限责任公司,安徽芜湖241000

出　　处：《工业仪表与自动化装置》2024年第5期83-87,共5页Industrial Instrumentation & Automation

基　　金：2021年安徽省高等学校质量工程重点项目(2021jyxm0105)。

摘　　要：为提高港口设备的远程检测和控制效率,该文基于高速移动通信和深度学习技术,设计了港口设备远程检测与控制平台。该平台采用B/S与C/S相结合的架构形式来提高平台响应速度,以独立的SA专网方式进行5G组网搭建,提高数据传输的安全性。通过融合残差结构与卷积神经网络,建立了具有信息有效传递的IDCNN模型,来提高数据样本数量少的特征提取准确度问题。测试表明,所搭建5G专网的SINR为13.98 dB,RSRP≥-85 dBm,能够满足平台任务需求。与SVM、CNN和决策树模型相比,所提模型的故障识别精度可达86.1%以上,证明了该方案的可行性。To improve the efficiency of remote detection and control of port equipment,this paper designs a remote detection and control platform for port equipment based on high-speed mobile communication and deep learning technology.The platform adopts a combination of B/S and C/S architecture to improve the corresponding speed of the platform,and builds a 5G network through an independent SA private network to improve the security of data transmission.By integrating residual structures and convolutional neural networks,an IDCNN model with effective information transmission was established to improve the accuracy of feature extraction with a small number of data samples.Tests have shown that the constructed 5G private network has a SINR of 13.98 dB and an RSRP of≥-85 dBm,which can meet the platform′s task requirements.Compared with SVM,CNN,and decision tree models,the proposed model achieves a fault recognition accuracy of over 86.1%,proving the feasibility of this scheme.

关 键 词：港口设备 远程检测与控制 高速移动通信 深度学习 卷积神经网络 

分 类 号：TN929.11[电子电信—通信与信息系统]