The Design And Research Of Campus Environment Monitoring System Based On Internet Of Things

Fire and environmental pollution accidents,has always been a serious threat to people's property safety and life and health,campus as a crowded place,once such accidents happen,it will cause immeasurable serious consequences.Most of the campus safety accidents are caused by the untimely monitoring and management,and the failure to take the best remedial measures in time after the accident.Therefore,effective monitoring and management of the campus environment has important practical significance.At present,most of the environmental monitoring systems are based on Zig Bee,WLAN or operators' 4G connection mode.This kind of connection mode has the defects of short communication distance,low capacity and high cost,which can not meet the needs of low power consumption,long distance,high capacity and low cost of the Internet of things now and in the future.In addition,this kind of environmental monitoring system has low degree of visualization of data,and lacks of subsequent data processing and analysis.In view of the shortcomings of the current environmental monitoring system,this paper designs a campus environmental monitoring system based on LoRa wireless network.The specific research contents are as follows:(1)The LoRa terminal node is designed and built,including hardware design and software design.The hardware design part includes the schematic design of the terminal module,PCB drawing,sensor and wireless module selection;the software part includes the terminal embedded software design and gateway network settings.The communication distance and power consumption of LoRa terminal node are tested.The test results show that LoRa communication mode can realize the long-distance communication within 500 meters of campus,and the power consumption is low.(2)Based on the framework of Things Board,the back-end system of campus environmental monitoring is designed and built.The main functions include user management,node visualization management,data visualization display and abnormal alarm.All kinds of software functions of the system are tested.The test results show that the system can visually monitor the environment values in real time,and provide web page and email alarm functions for abnormal data.(3)A CNN-BLSTM algorithm model based on attention mechanism is designed and built for ambient temperature prediction.All kinds of traditional environmental monitoring systems collect real-time data for monitoring,and there is no in-depth processing and application of the collected data.In order to further strengthen the analysis and utilization of data,aiming at the characteristics of the existing data sources,this paper first builds the LSTM algorithm model,and carries out two levels of optimization design on the basis of the algorithm model.Firstly,aiming at the shortcomings of LSTM model,such as lack of processing individual abnormal data interference and only forward training to mine data connection,CNN and blstm are introduced to extract temperature trend and pre and post features respectively.Secondly,the attention mechanism is introduced to optimize and reconstruct the model according to the characteristics of different environmental features.Finally,by comparing with FNN,RNN,LSTM and cnn-blstm time series models,it is concluded that cnn-blstm time series prediction model based on attention mechanism has better ability to predict future temperature data.