Research On Atmospheric Environment Key Rules Mining Based On Apriori Algorithm And Evidence Reasoning

Air pollution is one of the main threats to human health.How to control air pol-lution has become the focus of research and the society.The grid monitoring system has exerted positive effects for the treatment of air pollution by providing comprehen-sive information.Based on its monitoring data,single source monitoring data mining and multi-source data direct fusion mining methods have been proposed to explain the inter-attribute and mutual influence.Existing methods do not fully consider the uneven distribution of environmental monitoring data and the characteristics of climate change,so the reflection on cause of pollution is not comprehensive and accurate.Aiming at the limitations of existing methods,this paper studies the key rules mining methods and tech-niques of atmospheric environment from the perspective of data mining and uncertainty information fusion theory.The main research contents are as follows:(1)A key rule mining method for atmospheric environment monitoring data based on Apriori algorithm and Dempster-Shafer theory is proposed.Firstly,the Apriori algorithm is used to mine the association rules of the pre-processed single monitoring station data to obtain high-confidence strong association rules.Then,the above-mentioned strong association rules are merged by Dempster-Shafer theory to obtain hidden related rules from multiple monitoring stations data.The method can solve the data differentiation problem existing in the data mining process of multiple monitoring sites,that is suitable for the fusion requirements of diversified rule mining requirements and rule sets.(2)A key rule mining method for atmospheric environment monitoring data based on Apriori algorithm and Evidential Reasoning algorithm is proposed.Firstly,the weight index system is introduced.According to the importance of rule sets in fusion process,the corresponding weights of the strong association rule set extracted by Apriori algorithm are given.Then,the Evidential Reasoning algorithm is used to fuse multiple rule sets to obtain high confidence fusion rule set.The method can solve the problem of non-uniformity and partial rule conflict of the fusion data set under special circumstances,and that is suitable for high-precision mining and fusion requirements with different importance.The mining results of the method presented in this paper represent the influence mode between different parameters,which has interpretability and practical significance,and provides theoretical and technical support for air pollution control and prevention.