Study On Comprehensive Evaluation Model Of Air Quality In Xining City

The information mining of air quality monitoring data in Xining City has great significance for the understanding of air quality and the comprehensive treatment of air pollution.This article is published in the national"China Environmental Air Online Monitoring and Analysis Platform Xining Ambient Air Quality Index?AQI?,Weighted Euclidean Distance Comprehensive Evaluation Index?AQI'?,sulfur dioxide?SO₂?,and nitrogen dioxide?NO₂?inhalable particulate matter(PM₁₀),carbon monoxide?CO?,ozone?O₃?,and fine particle(PM_2.5)concentrations of six indicators were used as research objects.The following studies were conducted:1)Through the China Air Monitoring and Analysis Platform for the Environment,it collected the air quality monitoring index data of the government for a certain month in Xining city and calculated the air quality index AQI value and weighted Euclidean distance comprehensive evaluation index AQI';2)The trend graphs of the indicators of the day,week and month,and the trends of the AQI and AQI'are plotted and analyzed;mainly based on the air quality indicators AQI of the air quality monitoring and calculations.The weighted Euclidean distance comprehensive evaluation index AQI'and the six index concentration data of SO₂,NO₂,PM₁₀,CO,O₃,PM_2.5.5 were analyzed.Diurnal,weekly,and monthly changes in air pollutant concentrations were analyzed,and AQI'and AQI were compared.Trend graphs showed that AQI'values were slightly lower than AQI values,and the results were in line with actual conditions.3)Based on the qualitative analysis of the regularity of the monthly variation trend,the main pollution indicators of the six monitoring indicators are also quantitatively analyzed using principal component analysis,factor analysis and cluster analysis.Use these analytical methods to obtain air quality indices and verify them;4)Using a weighted Euclidean distance comprehensive evaluation index to evaluate air quality and establish a mathematical model that can quickly map the relationship from x?monitoring indicator concentration?to y?air quality index?,thus allowing decision makers to abstract the standard Different indicator concentrations have an intuitive,easy-to-comparison ranking index.A multiple linear regression model with a goodness of fit of 99.6%was obtained by multiple linear regression:y=0.084x₁+0.021x₂+0.157x₃-0.160x₄+0.168x₅+31.943Through the correlation coefficient R²evaluation,F-test,P-test,residual analysis and comparison of observations and predicted values,it is concluded that the multiple linear regression model is well fitted to the data and the accuracy is high,indicating that the model is fully available.