Research On Monitoring Data Analysis And Forecasting Methods Of Air Quality

This applied research focuses on the air quality problems in the construction of "National Central City" and the Free Trade Port.According to the daily air quality monitoring data of 10 cities from January 2014 to February 2023,Beijing,Tianjin,Shanghai,Guangzhou,Chongqing,Chengdu,Wuhan,Xi ’an,Zhengzhou and Haikou,the air quality monitoring data analysis and prediction model research were carried out.In terms of the analysis of air quality monitoring data,this application research firstly establishs two kinds of visual city AQI calendar charts,which can help us quickly understand the air quality level of a city every day in a year and the change regulation of AQI index in a year.Secondly,this application research used Hierarchical Clustering to divide 10 cities into 5 categories according to the concentrations of six basic pollutants,and the concentrations of six basic pollutants in each category were similar.Descriptive statistical analysis of AQI change and pollutant concentration was carried out,the correlation between the concentrations of six basic pollutants and AQI in each city is obtained,as well as the correlation between the concentrations of six basic pollutants.In terms of the study on air quality prediction methods,this application research study the long-term prediction of AQI,and establish the ARIMA prediction model which can automatically determine the order.By using deep learning methods,Bi LSTM model and CNN-Bi LSTM model are established.By comparing the prediction results of ARIMA model,Bi LSTM model and CNN-Bi LSTM model,this application research found that the CNN-Bi LSTM model has the best prediction effect,CNN-Bi LSTM can accurately predict the change trend of long-term air quality,which helps to make the time arrangement of large-scale outdoor activities and prepare for air pollution in advance.By accurately predicting the change trend of long-term air quality,this application study is helpful for relevant departments to make the time arrangement of large-scale outdoor activities and prepare for air pollution in advance.