基于数据驱动模型的大气污染扩散路径预测研究

打开文本图片集

文章编号：1674-6139（2025）12-0142-06

中图分类号：X831文献标志码：B

Abstract：Toachievehigheraccuracyinairpolutionanalysis，thisstudydesignedadata-drivenmodelbasedmethdforpredicting thedifusionpathofairpolution.ItusesGaussanplumemodeltosimulatethedifusionofatmospericpollutantsfromexprimental pointourcesinamicmospericeviroentndoainssiulateddiusiononentratioalussingirtalumadal hicles.Usingparticlefilteringsdtailationetdfordatarivemodelingtepaperintegatesoevatiositotussi anplumemodeltodynamicallstimatethestateparametersofamosphericpollutantdifusionsystemsandobtainpredictedconcetration distributions.The test found that thepredicted PM2，5 diffusion path obtained by this method is close to the actual diffusion path.Compared to PM2.5 ，the difference in prediction results for PM10 is smaller because PM2，5 has a smaller particle size and is more susceptible to the influence of airflow，making prediction more difficult.

Keywords：data-drivenmodel；particlefileringmetod；difusionofatmospericpution；Gaussianplumemodel；pathprediction

前言

大气污染扩散路径预测研究在当前全球环境问题日益严峻的背景下显得尤为重要。（剩余5361字）