Design And Development Of Real-time Environment Monitoring System For Forest Fire Engine Based On IoT

As the society developing rapidly,we begin to pay more and more attention to ecological civilization.Aiming at forecasting of forest fire and auxiliary command of forest fire suppression,this paper designed a system fixed on a forest fire engine to collect the real-time environment information around the fire engine.It realizes the real-time monitoring and transmission of wind speed,wind direction,air temperature,air humility,C02 concentration,CO concentration,smoke concentration,atmospheric pressure around the fire engine and the geographic location information.Using this data to build an simple fire identification model.The following content is the main things we had done:(1)On the basis of analyzing the current situation of forest fire monitoring methods and monitoring communication and fire command communication,a design of real-time monitoring system for forest fire engine based on satellite communication was put forward.(2)Using the BDS developed in China to achieve the positioning of the system and the transmission of information.Using two sets of wind speed and wind direction sensors and speed sensor,a correction model of 3D space vehicle attitude,we designed a system that could measure the horizontal wind speed and direction in any situation.(3)Complete the hardware design and the related software design.The hardware design mainly includes the schematic design of core control circuit,power module,BDS communication module and sensor acquisition circuit.The software part includes the main program of STM32 MCU based on C language,the compilation and debugging of each module program,the compilation of text storage program based on Fatfs and the design of management software of host computer.(4)Designe a series of experiments,such as fire simulation test,static wind speed experiment,communication reliability test and system overall performance test.The results shows that the system could collect the information accurately and the communication is reliable.(5)Obtaine data with our system,and pretreated the data.Then we use the BP neural network and SVM to build models for fire identification.Compared with SVM,BP neural network has higher fire correct recognition rate,reaching 98.67%.Finally we complete the software generation and release of the detection system,and made a GUI interface.