Design And Application Of A Data-driven Highway Tunnel Operation Monitoring System

Tunnel safety is an important content of highway operation and management,especially for the monitoring of fire condition and traffic flow state,the effect of various monitoring methods are still not satisfactory.This paper presents a data-driven highway tunnel operation monitoring system design method,and studies four aspects,including the monitoring data acquisition and processing,tunnel abnormal condition discrimination,monitoring linkage and monitoring plans generation,and monitoring data management.(1)Tunnel monitoring data acquisition and processingAnalyzing the purpose of tunnel monitoring system and characteristics of tunnel abnormal conditions,selection criteria of the monitoring parameters are proposed,and the tunnel environment and traffic condition parameters are selected respectively.The data sharing of tunnel operation monitoring is realized by using cloud computing technology,and the data preprocessing and data fusion method are verified by experiment simulation,which provides effective data support for tunnel safety situation evaluation.(2)Tunnel abnormal condition discrimination and safety situation assessmentThe common abnormal conditions and its influence of highway tunnels are analyzed,and priori knowledge models for fire and traffic jam condition discrimination are put forward,by training the cleansed tunnel monitoring data with supervised learning method.A tunnel safety evaluation index system and a BP neural network algorithm based tunnel safety situation assessment method are proposed,and the effectiveness and accuracy of the discrimination model and the situation assessment method are verified by examples.(3)Tunnel monitoring linkage and monitoring plans generationThe common abnormal conditions and its influence of highway tunnels are analyzed,and priori knowledge models for fire and traffic jam condition discrimination are put forward,by training the cleansed tunnel monitoring data with supervised learning method.A tunnel safety evaluation index system and a BP neural network algorithm based tunnel safety situation assessment method are proposed,and the effectiveness and accuracy of the discrimination model and the situation assessment method are verified by examples.(4)Tunnel operation monitoring data managementBased on the analysis of the functional requirements of the tunnel operation,the overall structure of the monitoring data management system and its function modules are designed,including the equipment layer,the equipment control layer,the communication layer,the storage layer,the data processing layer and the monitoring management layer.A B/S-based highway tunnel operation monitoring data management system is designed,and the system interfaces are shown.