Research And Implementation Of Home Fire Monitoring System Based On Internet Of Things Cloud Platform

The living style of residents in China is mainly high-rise buildings.Due to the relative concentration of personnel and small space in high-rise buildings,in case of fire,the life safety and economic property of residents will face a serious test.How to quickly identify and eliminate fire in the early stage of fire has become an important research problem at present.Aiming at the actual home environment,this paper studies a smart home fire monitoring system based on the Internet of things cloud platform,combined with multi-sensor data fusion technology and flame image recognition technology,in order to realize the real-time monitoring of fire data,improve the accuracy of fire identification,and solve the problems of real-time and reliability of the system.The main research work of this paper is as follows:(1)This paper studies the function,resource model and communication protocol of onenet Internet of things cloud platform,and studies and analyzes the working principle and implementation method of related technologies involved in the system,so as to lay a foundation for the development and design of the follow-up system.(2)Aiming at the problem that the single sensor system is easy to lead to false alarm and missing alarm of fire,it is proposed to take STM32 as the core processor to build a multi-sensor hardware acquisition system to collect the data of multiple fire factors through the analysis of the electrical characteristics,working principle and circuit principle of modules such as temperature and humidity,flame,smoke and camera,combined with esp8266 module and relay module,Realize wireless communication with onenet cloud platform and control of remote devices.(3)The initialization program,data acquisition program,communication program and relay control program of each hardware module are designed based on STM32 firmware library;Based on EDP protocol,the data encapsulation and analysis program of sensor;Based on the deployment of onenet cloud platform,the functions of data storage,monitoring and early warning are realized,and the visual display of data and the issuance of control commands are realized by UI application controls.(4)Aiming at the reliability problems of the system,an improved fusion algorithm based on DS evidence theory is proposed to solve the problems of false alarm and missing alarm of fire caused by sensor fault or noise interference.In the application of DS evidence theory in fire system,the selection of fire characteristic quantity,multi-sensor data normalization finally,through the analysis of an example,it is proved that the improved multi-sensor data fusion algorithm based on DS evidence theory is effective to improve the accuracy,stability and reliability of the system.(5)Aiming at the problem of insufficient real-time performance of the system,a technical solution of flame image recognition is proposed.Firstly,the suspected flame area is segmented by using the inter frame difference method,it is found that the inter frame difference method has the problem of weak anti-interference when processing the scene of multiple moving objects,combined with the flame feature recognition algorithm to further process the image,it is still unable to distinguish the light sources with similar features.Therefore,an algorithm based on Ada Boost cascade classifier is proposed to improve the accuracy and timeliness of flame recognition through machine learning training.(6)The system test shows that the overall function of the system runs normally,and the reliability,real-time and practicability of the system achieve the expected goal.