基于决策树的煤化工污染物定量化溯源研究

QUANTITATIVE TRACEABILITY OF COAL CHEMICAL POLLUTANTS BASED ON DECISION TREE

环境污染问题日益严峻,煤化工生产是环境污染的重要源头之一,但由于煤化工生产过程中所产生的污染物种类繁多且各类产品的生产量不同,污染问题通常难以定量化溯源。挖掘环境污染物的产生原因、分析研究不同产品的生产和产量对各类环境污染物(如PM2.5、SO2)的量化影响是有效解决环境污染问题的必要前提。数据挖掘方法能够挖掘出隐藏在数据中知识或关联关系,其中的决策树算法以香农的信息论为理论背景,能计算出多个影响因子中影响力的大小关系,并且还能找到数量型属性的临界点,是环境污染定量化溯源的有效方法。本文利用决策树算法构建环境污染物量化分析模型,以宁东能源化工基地为例,研究解析煤化工产品的生产状况对环境污染物含量的影响情况。

The environmental pollution is increasingly servere,and the coal chemical industry is one of the momentous sources of environmental pollution. Since the kinds of pollutants generated in the production process of coal chemical industry are various and the throughputs of all kinds of products are different,it is usually difficult to quantify traces of the source of the pollution problem. It is a necessary precondition for the effective solution to the environmental pollution problem to study the influence of the production and throughputs of different products on the environmental pollutants( such as PM2. 5and SO2).Data mining methods can dig out the knowledge or relationship hidden in the big data. The decision tree algorithm from Shannon’s information theory can calculate the size relationship of the force of many influence factors,and find the critical points of quantitative attributes. The algorithm an effective method to quantify traces of the source of the pollution of the environment. In this paper,the decision tree algorithm is used to build the quantitative analysis model of environmental pollutants. In order to study the influence of the production situation of coal chemical products on the environmental pollutant content,as a case study,Ningdong Energy and Chemical Industry Base is taked as an example.