Research On Key Transport Path Mining Methods Of Air Pollutants

In recent years,China's industrial development in many areas has led to the deterioration of air quality.In order to establish an effective regional air quality monitoring system,it is particularly important to excavate the critical transport path of air pollutants,which has an important impact on air quality prediction.Based on the above purpose,this paper proposes a method to construct a network model of air pollutant transport.On the basis of this model,clustering algorithm is applied to mine the critical transport path of air pollutants,reveal the transport law of air pollutants,and take effective pollution prevention and control measures to improve air quality.Firstly,in order to explore the transport law of air pollutants,this paper analyzed the temporal and spatial distribution changes of air quality,found the main factors affecting the spread of pollutants,extracted the main transport characteristics of pollutants,and provided theoretical support for this paper to excavate the critical transport path of pollutants and verify the validity of the critical transport path.Secondly,based on the main transport characteristics of pollutants,a measurement index reflecting the local pollutant transport ability is established,which describes the local pollutant transport relationship,constructs the directed weighted complex network model of air pollutant transport,abstracts the monitoring station as a network node,and the transport relationship as a network path.Meteorological and geographical factors determine the complex network path miscellaneous.Thirdly,based on the established pollutant transport network model,a clustering algorithm for mining the critical pollutant transport path is established.Aiming at the problem that the clustering granularity of traditional AP clustering algorithm is affected by biased parameters,an improved IAP clustering algorithm is proposed,which can reasonably control the output of clustering granularity by introducing standard mutual information between transport paths.To solve the problem of slow execution of traditional AP clustering algorithm,an improved CFAP clustering algorithm is proposed based on IAP algorithm and Concurrent.Futures parallel development framework.Line optimizationimproves the execution speed of the algorithm.According to the results of cluster analysis,outliers from one group were selected,and the other cluster centers were the key pollutant transport paths.This critical transport path is representative in the case of guaranteeing obvious local transport relationship.Finally,the experimental analysis of the method proposed in this paper verifies that the air pollutant transport network constructed has obvious scale-free characteristics of complex network;through the evaluation and analysis of the key air pollutant transport path obtained by mining,the effectiveness of the proposed method is verified,and the key propagation path obtained by mining is reasonable.