Research And Implementation Of Dynamic Monitoring Technology For Highway ETC Gantry Health

Since Premier Keqiang Li’s request at the National People’s Congress in 2019,China has conducted a large-scale nationwide deployment and networking of the electronic non-stop tolling ETC system for motorways,and the ETC system has become one of the largest Internet of Things systems in the world.Different from the traditional regional Io T systems,ETC gantry systems suffer from common industry problems such as multiple types of equipment,wide distribution,lack of experience in equipment operation and maintenance,and the equipment requires 24-hour uninterrupted operation in various harsh outdoor environments.At present,the monitoring efficiency of the health status of ETC equipment and systems by provincial expressway companies is insufficient,and it is difficult to effectively achieve dynamic tracking and management of faults.Based on the mining and analysis of ETC transaction data,this paper conducts in-depth research on the health status monitoring of the operation of the ETC gantry system of provincial expressways.The specific research work mainly includes:Ⅰ.Firstly,the basic attributes and characteristics of ETC big data are deeply studied,and different strategies and methods are designed and used to clean and correct the problems such as data errors and data redundancy in ETC big data."The gantry position information correction method for vehicle trajectory data and ETC big data prepares the data foundation for further research.Ⅱ.The highway network contains rich road network topology information and is continuously updated.How to quickly and effectively extract and update the highway path set for each trip start OD pair has become a key problem to be solved urgently in ETC big data verification.This paper proposes a road network model construction method suitable for the structural characteristics of the expressway network.Based on the highway network topology,the node matrix of the toll station and ETC gantry nodes is constructed,and then the Dijkstra algorithm is used to construct the path gantry for each OD pair.The sequence set can effectively support the rapid extraction of trajectory features of ETC big data.Ⅲ.A DTW algorithm with dynamic search steps is designed to achieve fast detectionof missed and erroneous transactions in ETC big data by performing similarity metrics and anomaly addressing on the ETC data trajectory and path gantry sequences of each OD pair.Experimental results show that the model has a detection accuracy of over 99%,while improving retrieval time efficiency by more than 7 times and 10 times compared to the industry’s mainstream algorithms DTW,Hausdorff and EDR.Ⅳ.Based on the above work,a dynamic monitoring method for ETC gantry health based on big data is designed.Firstly,Kafka and Flink technologies were used to build an ETC gantry health monitoring model,and then a dynamic monitoring system for ETC gantry health in a province was designed through front-end Web technology to achieve dynamic monitoring of ETC gantry health status and improve the intelligence level of ETC gantry system operation and maintenance.