Real-time Monitoring System Of Height Limiting Device Based On Internet Of Things

As the bumper of the bridge,culvert and interchange,the height-limited protective frame can serve the purpose of warning identification and protection of the bridge,culvert and interchange.At present,the maintenance and detection methods for overhead limits mainly rely on periodic inspection by personnel,which is time-consuming and laborious.At the same time,some areas adopt the high limit monitoring scheme of laser and infrared,which has not been widely popularized.The main reason is that it is difficult to set up and debug,and it is greatly affected by topographic factors.There are many interference signals and the system has a high false alarm rate.The interception of super-high vehicles is not mandatory and the protection of bridges and culverts is limited in practice.Using the "cloud-edge-end" architecture,this paper develops a real-time monitoring system for height-limiting devices based on the Internet of Things.The main research contents are as follows:(1)The system is composed of vibration monitoring subsystem,image information collection subsystem and cloud service subsystem.Among them,the vibration monitoring subsystem uses the low power consumption triaxial accelerometer ADXL362 as the main method of vibration information monitoring.MSP430 and NB-Io T modules are selected as the information processing module and network transmission module,and their circuits are developed and designed.The low power consumption devices are selected to ensure the low power consumption performance of the system.Reasonable circuit design further improves the system stability.The image information acquisition subsystem uses 4G network cameras to support secondary development.At the same time,a photovoltaic power supply system is designed for the subsystem,which meets the design concept of green construction.(2)The EMD decomposition algorithm is used to filter the received acceleration data,which can effectively reduce noise and improve the accuracy of the signal.The dynamic thresholds of the data are determined and the vibration behavior is effectively divided.Acceleration data are collected regularly for posture calculation,and the static roll angle and pitch angle of the current equipment are calculated.Vibration signal reception service program based on C#language,image acquisition and forwarding service program,SMS alarm service program and so on are developed with VS2019 development tool to realize the interactive transmission of data among the three subsystems.Design database tables and triggers based on SQL SERVER to ensure data integrity and consistency;The vehicle information recognition technology based on PP-YOLO is studied.K-Means++ algorithm is applied to re-cluster the training dataset,and good fitting results are obtained.(3)This paper designs and develops intuitive and convenient management pages,twodimensional and three-dimensional visualization large screen,We Chat applet client,etc.According to the site environment provided by Tianjin Huatiekewei for Technology Co.,Ltd.and Beijing Railway Administration,the system can run stably after actual collection and test,and display data normally in Web client and We Chat applet client,which verifies the feasibility of the system.