| In recent years,the problem of air pollution has become an unavoidable reality for Chinese cities.Among all air pollutants,PM2.5 pollution is the most serious and causes the most damage to human health.PM2.5 pollution is becoming increasingly serious,and pollution control is urgently needed.In addition,China has entered the "new normal" and the economic situation is not optimistic,which has also increased the difficulty of governance.Many domestic and foreign scholars have done a lot of research on the treatment of PM2.5 pollution,mainly studying the treatment of PM2.5 pollution in Beijing,Shanghai,Guangzhou and other first-tier cities,and proposing the treatment methods of PM2.5 pollution.In the current economic environment,it is an urgent problem for the current government to effectively control PM2.5 pollution while maintaining steady economic development.Among large and medium-sized cities,Chengdu has the most serious PM2.5 pollution.This paper intends to study the treatment of PM2.5 pollution in Chengdu on the basis of existing research results.Compared with the existing research literature,in order to adapt to the current economic environment and effectively control PM2.5,it is necessary to actively explore the combination of long-term and short-term treatment measures.The main research contents are as follows.PM2.5 pollution is a long-term problem that cannot be fundamentally solved in a short period of time.Quantitative analysis of the influencing factors of PM2.5 pollution can carry out long-term planning and treatment of PM2.5 pollution under the condition of determining the specific environmental situation.Considering the two-way interaction between regional pollution and regional economic growth,simultaneous equations model was selected to quantitatively analyze the influence of various economic factors on PM2.5 concentration,and long-term treatment measures were proposed based on the regression results.Long-term measures can be taken to control air pollution at the source,but the problem of ecological environment deterioration is becoming more and more serious.Moreover,air pollution control cannot be accomplished overnight.Large-scale and simple closure of enterprises with high pollution and high emissions will inevitably improve the environmental quality,but also lead to a cliff-edge economic decline.Therefore,under the long-term control measures of PM2.5 pollution,short-term pollution prevention and control should be combined.Short-term pollution control and emission reduction before the arrival of heavy pollution weather to avoid heavy pollution weather,which puts forward requirements for the accuracy of PM2.5 concentration prediction in the future.By referring to relevant literatures,the machine learning method has unique advantages.The circular neural network structure with attention mechanism is trained to predict the PM2.5 concentration in the next 7 days,which can be used to issue an early warning before the arrival of heavy pollution weather and provide policy guidance for the government to take emission reduction measures.The following main conclusions are obtained.First,a comparative analysis of PM2.5 pollution in Beijing,Shanghai,Guangzhou and Chengdu shows that the average duration of pollution in Chengdu is 6.85 days,much higher than the other three cities,with the most serious pollution.At the significance level of 5%,whether there is a significant decrease in PM2.5 concentration in 2018 compared with the same period in previous years is tested.The results showed no significant improvement in air pollution,with significant increases in pollution in some periods,and the treatment of PM2.5 pollution in Chengdu is extremely urgent.Secondly,the simultaneous equations model was used to study the long-term influencing factors of PM2.5 pollution in a unified framework,so as to control air pollution at the source.The model results show that dust from coal and construction sites is the main cause of PM2.5 pollution.To completely improve air quality,the government should advocate the use of clean energy in the whole society,prevent centralized construction in the short term,and improve citizens' awareness of environmental protection.Taking short-term emission reduction measures before the arrival of heavy pollution weather can effectively reduce the concentration of short-term pollutants and achieve short-term treatment.Short-term PM2.5 concentration is affected by meteorological conditions.In order to improve the prediction accuracy and control short-term PM2.5 pollution more scientifically,it is necessary to analyze the influence of meteorological conditions on PM2.5 concentration.The results showed that under the conditions of static wind,the air flow is weakened and the pollutants are not easy to spread,resulting in short-term PM2.5 pollution.Finally,based on the historical information of meteorological conditions and pollutant concentration,the circular neural network with attention mechanism was trained to predict the PM2.5 concentration in the next 7 days.The test set was selected as July and August 2018,and the root mean square error was 10.71.When the model predicts that the pollutant concentration will exceed the standard in the next 7 days,the government can take actions in advance to reduce emissions and slow down the heavy pollution.However,this paper also has some shortcomings.The influence of meteorological conditions on PM2.5 concentration in Chengdu is analyzed,and the results show that the proportion of southwest wind in the initial state of pollution increases,which indicates that the pollutants from southwest are transmitted to Chengdu along the air flow track,which is also the cause of local air pollution.In the face of severe air pollution,in order to effectively control,not only local emissions should be reduced,but also surrounding cities are important causes of pollution.In order to effectively control the local air pollution in Chengdu,more accurate policy support should be made by combining the current situation of pollution in surrounding cities. |
PDF Full Text Request