Research And Implementation Of Indoor Air Harmful Substance Monitoring System Based On Neural Network

With the continuous improvement of the level of industrialization and urbanization in China,environmental pollution is becoming more and more serious and presents a trend of diversification.Especially,indoor air pollution has brought great harm to people's health.As the people's pursuit of healthy living environment is getting higher and higher,the rigid demand for monitoring of air harmful substances is increasing,and the existing environmental monitoring devices are relatively fixed,not easy to expand,and can only be viewed as parameters for people to watch,and the function of data management is weak,and it is becoming more and more unsuitable for people's needs.Therefore,it is very necessary to design a new type of air harmful monitoring device.There are two reasons.Firstly,the demand for the collection of air harmful parameters is different on different occasions,and the monitoring device is required on the basis of good centralized management and communication ability,and the requirement for parameter monitoring in different environment can be carried out.Secondly,the monitoring device requires not only to collect and display data,or as the generation and provider of large data,but also as a data analysis and predictor,and can analyze and predict the trend of environmental parameters based on real time and historical data,and optimize the operating conditions of the measures.The main research work of this paper is as follows:Firstly,a kind of combined indoor air harmful monitoring device is designed,which adopts flexible and changeable inserting structure.The structure is divided into the bottom layer,the middle layer and the top layer.The function of the underlying module is to manage the middle layer module and the top layer module,communicate with the server in the public network by Wifi.The function of the middle layer module is to have the function of environment detection.Each middle layer module is responsible for the collection and transmission of a kind of air harmful parameters.The limit of the order and type,the number of the middle layer module,determines the variety of the monitoring parameters of the device;the optional display module on the top layer is located at the upper end of the middle layer module,and uses a good man-machine interface to display the real-time monitoring value of the environment parameters.Inter module communication conforms to the Modbus specification and transfers data to the server-side.Secondly,establish server-side to communicate with indoor air harmful environmental monitoring device through Ethernet.The server side technology implementation is divided into interface software,database and WeChat public number.The function of the interface software is to connect the data uploaded by the hardware device to the server.The function of the database is to store and manage the air harmful data collected by the monitoring device.After the analysis and processing of a large number of environmental parameters,the real-time and historical data and the trend prediction node are graphically displayed through the WeChat Subscription.Fruit,and alarm.Finally,for the prediction of the trend of indoor air harmful substances,an adaptive network-based fuzzy inference system(ANFIS)is used and run on the lower machine.Using the actual parameters collected by the device as a sample,the data is preprocessed and trained to determine the optimal structure of ANFIS.Several groups of PM2.5 and CO samples in the actual situation are used in ANFIS and Generalized Radial Basis Function Neural Network(GRBFNN)respectively.The back-propagation neural network(BPNN)model was tested and compared to prove the accuracy of the ANFIS model in predicting the trend of harmful indoor air.Through the application of this system,many kinds of indoor air harmful parameters can be collected and displayed,the change trend is predicted,and DDC can be driven by Modbus output prediction results to realize the linkage of equipment.It is of great practical significance to correctly predict the changing rules of harmful gases in indoor environment and improve the environment in real time.