Design And Implementation Of Air Quality Monitoring And Forecasting And Early Warning Software

The design requirements of this software are based on Shaanxi Province Science and Technology Co-ordination Innovation Project,project name is:"Internet+"remote mineral resources mining environmental pollution monitoring system?the project number is2016KTCQ01-26?,according to sub-project decomposition of the project,this paper develops a monitoring and forecasting software for air pollution.The software uses JAVA language,MVC architecture to build,and uses Android operating system and Baidu MTC test platform to complete the software framework,functional design and software testing.In addition,according to the environmental monitoring data of Xi'an,this paper builds a prediction model based on BP neural network and wavelet neural network time series algorithm,predicts the daily concentration and hourly concentration of atmospheric pollutants(SO₂,O₃,PM_2.5,PM₁₀)on the MATLAB platform,And further use of Android software to achieve the forecast data display.The main work accomplished in this paper are as follows:?1?Based on the actual functional requirements of group projects and the development of the domestic and international air quality monitoring and forecasting early warning system,the overall framework of the air quality monitoring system and the R&D program of the Android client software are analyzed and presented.?2?The algorithm of BP neural network and wavelet neural network time series are studied.Based on the training set and the forecasting set of pollutant data and corresponding meteorological data of the monitoring stations in Xi'an from 2015 to 2017,on the MATLAB platform,this paper predicts the pollutants(SO₂,O₃,PM_2.5,PM₁₀)by BP neural network prediction model in the long-term and forecasts the pollutants(SO₂,O₃,PM_2.5,PM₁₀)by using the wavelet neural network time series forecasting model.The experimental results of the above model are as follows:The prediction accuracy of BP neural network models for SO₂,O₃,PM_2.5.5 and PM₁₀0 are respectively:83%,65%,80%,79%;the forecast accuracy of the wavelet neural network time series prediction model for SO₂,O₃,PM_2.5.5 and PM₁₀0 is:74%,70%,64%,63%.?3?Android software is designed to realize the functions of the national weather forecast query,the real-time display of the data of the own air quality monitoring station,the historical data query,the map marking,the prediction and display of the concentration of the pollutants by using the JAVA programming language,the MVC framework and Android platform.?4?The software running effect test,software deep compatibility test and software deep traversal test have been completed by using Baidu MTC software testing platform.The test results show that the software developed in this paper can meet the functional requirements of the project,and the interface is beautiful and the operation is smooth.The software passed the test prototype at 85%of the time,with software installation time of11.41s,better than the industry average of 19.99s,startup time?0.46s?,CPU usage?5.6%?and traffic consumption?224KB?are all better than the industry average.The design of this software has a certain reference value to the design of PC software in the field of atmospheric monitoring.