Research On Edge Computing Technology For Smart Fire Protection Scenarios

The continuous development of urbanization is accompanied by the continuous increase of urban population and the shortage of urban land resources,which promotes the development of urban buildings towards high altitudes,and the number of high-rise buildings gradually increases.Among them,the fire safety of buildings is a common problem faced by the world.With the rapid progress and development of information technologies such as modern computers,the Internet,and big data,it has become very important to apply information technology to the construction of smart fire protection systems.Therefore,this thesis provides reliable support for fire safety by monitoring the operation status of fire protection facilities in real-time and grasping the status of the facilities in time.The rapid increase in the amount of data generated by Io T devices in the era of big data poses a huge challenge in response delay and network load.Through the combination of intelligent monitoring and edge computing technology,the monitoring data is transmitted to the edge for processing,which further improves the monitoring efficiency and relieves the pressure caused by the capacity and real-time transmission of the Internet.This thesis studies the edge problem in fire protection scenarios.By deploying edge devices close to the data collection end,data processing is completed and the results are returned to the server,to relieve the pressure of data communication in fire monitoring scenarios.The research contents and achievements of the thesis are as follows:1.The identification of the switch status of fire shutter doors of common fire protection facilities in engineering buildings is studied,and a real-time detection algorithm of fire shutter door switch status based on video and image processing technology is proposed.By collecting video images of rolling doors in different experimental scenarios,the overall characteristics of the images are analyzed,and the image preprocessing technology,edge detection,Hough line,and other methods are studied,to realize the effective identification of the opening and closing ratio of different types of fire shutter doors.Based on the proposed scheme,the software,and hardware implementation of the algorithm on the edge device Jetson Xavier NX is completed,and experiments and functional verification of the developed device are carried out in several typical scenarios.2.For the edge computing scenario of multi-channel data access,study the optimization strategy of resource allocation and unloading parameters of edge devices.Using queuing theory to model the service access delay and cost of multi-channel terminal data access,a computing resource allocation scheme is proposed,and a Delay-Price Equalization Algorithm Based on Interior Point Method(DPEAIPM),to solve multiple parameters such as unloading probability in the optimization problem,to realize the optimal resource allocation and unloading parameter optimization under the requirement of service quality delay.