Analysis And Prediction Of Air Quality In Beijing

In recent years,the problem of air pollution is becoming more and more prominent,and the condition of air quality is worrying,so it is very important to study air quality.Beijing is the capital of China and the center of national politics,culture and international communication.Air quality in Beijing is not only related to people' s health,but also to the image of the capital and the country.Therefore,it is significant to study the air quality of the city.This paper collates the daily air quality data of Beijing from 2016 to 2019,and the daily data includes AQI,air quality grade,PM2.5,PM10,SO2,CO,NO2 and 03.On the basis of the above data,this paper studies the air quality of Beijing from the following aspects:Firstly,descriptive statistical analysis of AQI and air quality gra-de are performed for each year.The results show that the mean and standard deviation of AQI are decreasing year by year.In each year,the proportion of good air quality grade is the largest,while the proportion of heavy pollution and severe pollution are small.From 2016 to 2018,the excellent and good rate increases year by year,and the excellent and good rate in 2019 is 1.6 percent lower than that in 2018.Secondly,the relationship between AQI and six pollutants is studied on the basis of the improved entropy weight-gray relational analysis method,so as to find out the main pollutant affecting the air quality of each season in Beijing.The results show that O3 is the main pollutant affecting air quality in summer,and the main pollutant affecting air qua-lity in spring,autumm and winter is PM10.Therefore,the prevention and control of O3 should be strengthened in summer,and the prevention and control of PM10 should be strengthened in remaining three seasons.Thirdly,in order to explore whether the air quality in Beijing has been continuously improved from 2016 to 2019,the annual mean of AQI and the excellent and good rate are used to rank the air quality in these four years.The results are different.In order to obtain more accurate and reasonable result,this paper adopts the improved TOPSIS method for fur-ther research based on the annual average concentration of six pollutants.The results show that the result obtained by the improved TOPSIS method is consistent with the result obtained by the excellent and good rate,and they both think that the ranking of air quality from good to bad is 2018,2019,2017 and 2016.Therefore,the above result is taken as the final ranking result of air quality.It can be seen that the air quality in the past four years has not showed a trend of getting better year by year,so we should continue to strengthen governance to make the air qua-lity continue to improve.Finally,the daily AQI data of the whole year of 2019 are selected to establish the ARMA(2,0)model,which is used to directly predict the AQI of the next five days,and then compare it with the result of rolling prediction.The result shows that the MAPE and MAE of direct predic-tion with ARMA(2,0)model are smaller than rolling prediction,that is,the prediction effect of the former is better than that of the latter.