基于粗糙-信息熵的水质综合评价方法

A WATER QUALITY COMPREHENSIVE ASSESSMENT METHOD BASED ON ROUGH-INFORMATION ENTROPY THEORY

水质综合评价是水环境治理的指导性工作,水质综合评价的信息量大,情况复杂多变,如何有效的提取有用信息以及客观地对各水质指标进行权重设定,一直是水质评价方法的核心难题。粗糙集理论是一种有效的处理不确定性信息的方法,能够在复杂的水质数据中挖掘有效信息。针对粗糙集理论实践中所存在的信息丢失、以及在系统有序化方面的不足,应用信息熵真实反映属性重要程度的特点对其进行修正,得到了一个比较完善的粗糙-信息熵水质综合评价方法。该方法继承了粗糙集和信息熵直接从数据出发进行信息挖掘的特点,较好地克服了传统水质评价方法主观性强的缺陷,同时还能够对同级水质进行细化分级,这对于污染治理、投入评估具有重要的借鉴意义。将粗糙-信息熵法和综合指数法、模糊数学法、层次分析法所得结果进行比较,结果证实该方法不仅评价过程简单明确,评价结果客观丰富,而且评价结果的精度和可靠性都有较大提高。其研究结论对水质监测数据的评价有一定的借鉴作用,亦对河流水污染防治对策的制定具有一定的指导意义。

With the development of social economy and the expansion of human activities, water quality has gradually deteriorated in many areas. Water pollution is a serious problem threatening the survival of human beings,plants and animals. Water quality comprehensive assessment, as a important reference for water environment treatment,is increasingly being valued by the people.Large amounts o~ information and complex situations are the radical character of water quality comprehensive assessment. How to effectively extract useful information and set weight of water quality index objectively,has long been a main problem of water quality assessment. With the ability to deal with both numeric and nominal information, and express knowledge in a rule-based form,the Rough Set Theory （RST） has been successfully employed in many fields such as image segmentation,location services,travel modeling, medicine and so on. However, the application of RST has not been widely investigated in water quality analysis.As a mathematical tool to deal with uncertainty information, RST can effectively exploit information from complex water quality data.Information entropy has an important application in RST,and it can solve the problem in information loss and resultant sort for RST,thanks to the characteristic which can reflect the importance of properties accurately.Based on rough-information entropy theory, a new method for assessing water quality is put forward in this paper. This method inherits the features of RST that can effectively exploit information from the raw data directly.For this reason,it can reduce the influence of the subjective factors which have a very large impact on the traditional water quality assessment methods. In addition, the method based on rough-information entropy theory can further make water classification for monitoring sections which are the same pollution levels, and it has important meaning to the estimation of input in funds and technology for water pollution control.By comparing the evaluation results with the ones obtained from comprehensive index method, fuzzy synthetic evaluation method and principal component analysis method, it was proved that the method established in this paper is not only simple and clear, but also rich and objective in the content. What＇s more, it can refine classification and make the evaluate results more exact and more reliable.The research can not only provide some certain references for water quality evaluation,but also can offer guidance and recommendation to the measures for water pollution control.