Analysis of Chengdu air quality pollution based on logistic sequential multi-classification of distance between classes
Vol 2, Issue 1, 2021
VIEWS - 46 (Abstract)
Abstract
Air quality is closely related to people’s health. It is very important to analyze the pollutants affecting air quality. The sequential Logistic multi-classification method of inter class distance was used to analyze the air quality data of Chengdu from May 2019 to April 2020. Based on the inter class distance, the multi classification problem was transformed into multiple binary classification problems, and then the binary classification Logistic was used based on the sequential principle. Finally, the correct rate after stepwise regression was used to analyze the pollutants affecting air quality. Experimental results show that the four types of pollutants PM2.5, PM10, NO2 and O3 have the greatest comprehensive impact on air quality of Chengdu. The government should strengthen joint monitoring of these types of pollutants and formulate corresponding policies to reduce pollutants.
Keywords
Full Text:
PDFReferences
1. Zhang Baolin, Li Jiannan, Zhao Junling, et al.Effects of aeolian dust on urban air quality in Inner Mongolia grassland[J]. Science Technology and Engineering, 2018, 18(3) : 123-131.
2. Liu Ling, Zhao Qiaohua, Wang Jing, et al.Analysis of a winter-time heavily polluted weather process in the yangtzeriver Delta urban agglomeration[J]. Science Technology and Engineering, 2019, 19(26) : 376-383.
3. Xu Mengchen, Zhang Lijun, Zheng Nan, et al.Analysis on the change trend and pollution characteristics of urban air quality after revision of ambient air quality standards: take Jinan City as an example[J]. Science Technology and Engineering, 2020, 20 (13) : 5422-5428.
4. Li Na, Wei Xin, Zhou Yufeng, et al.Source analysis and health risk assessment of polycyclic aromatic hydrocarbons in atmospheric environment PM2.5 in Changchun City[J]. Science Technology and Engineering, 2021, 21(1) : 410-416.
5. Xie Shangfeng, Zhou Zhihao, Li Guohong. Relationship between PM2.5 concentration and meteo-logical factors in Nanning[J]. Science Technology and Engineering, 2020, 20(1) : 460-466.
6. Zhang Xiaokai, Yu Lei, Sun Miaomiao, et al.Distribution and elemental analysis of particulate matter PM2.5, PM10 in the atmosphere of spring in Jinan City[J]. Science Technology and Engineering, 2018, 18(25) : 278-285.
7. Tian Haitao. Linear regression model estimation method with ordinal multi-category explanatory variables and its application research[D]. Chengdu: Southwest Jiaotong University, 2019.
8. Li Jinxiu. Multi-classification research based on Logistic regression model and support vector machine (SVM) model[D]. Wuhan: Central China Normal University, 2014.
9. Murphy P M, Ahad W.UCI repository of machine learning database[EB/ OL]. (2007-6-25)[2020-4-30]. Http: www.ies. Uei. Edu/~mlearn/ mlrespository. Html.
10. Wangyushi PM2.5 air quality data, historical data and meteorological data - data network [eb/ ol] (2013-3-21)[2020-4-30]. Http:// www.shu-ju. Net/.
11. Wuyicheng Weather report [eb/ ol] (2019-5-1)[2020-4-30]. Http:// www.tianqihoubao. Com/.
12. Chengdu Bureau of Statistics. Chengdu statistical yearbook 2018[M]. Beijing: China Statistics Press, 2018.
13. Urban Planning Newsletter. Baidu Maps released the“China Urban transportresearchreport for the Second Quarter of 2018”[J]. Urban Planning Newsletter, 2018(14) : 14.
14. Cao Yang, Wang Chenxi, Zhao Xiaoli, et al.Analysis of the characteristics of PM2.5 pollution in Chengdu and its relationship with surface meteorological elements[J]. Mid-low Latitude Mountain Meteorology, 2020, 44(4) : 59-64.
15. Li Polish, Luo Bin, Zhang Wei, et al.Analysis of temporal and spatial distribution characteristics and pollution characteristics of ozone in Sichuan Province[J]. Environmental Science and Technology, 2018, 41(S1) : 293-298.
16. Yang Xianyu, Yi Jiajun, LüYaqiong, et al.Analysis of the causes of severe ozone pollution in Chengdu and surrounding areas[J]. China Environmental Science, 2020, 40(5) : 2000-2009.
Refbacks
- There are currently no refbacks.
