Research on Association Rules Mining of Atmospheric Environment Monitoring Data

Abstract

Wireless sensor network technology and gird monitoring system have exerted positive effects on the treatment of air pollution by providing comprehensive information. Based on its data monitoring, multi-source data direct fusion mining methods have been proposed to explain the interaction among these data. Existing methods did not fully consider the uneven distribution of environmental monitoring data and the characteristics of climate change. This paper studies the association rules mining methods and techniques of the atmospheric environment from the perspective of data mining and uncertainty information fusion theory. The association rules mining atmospheric environment monitoring data method are proposed based on Apriori algorithm and Dempster-Shafer theory together with Apriori algorithm and Evidential Reasoning algorithm. In this paper, the experiments of different fusion sequences are carried out using the data of China national control stations and USA micro stations, which are divided into the order of the first time, then space and the first space, then time. The time fusion is to fuse the rules of different monitoring stations in the same month. The spatial fusion is to fuse the rules of different monitoring stations in the same time range. The experiment changes the order of fusion to get the association rules between pollutants. Mining result representation of the mode of influence between different parameters is interpretable and practical, providing theoretical and technical support for the treatment and prevention of air pollution.