基于组合赋权-属性识别法的水质评价

摘要

针对传统赋权的水质综合评价模型中存在评价因子间联系不够紧密的缺陷,提出组合赋权思想.引入最优传递矩阵概念改进层次分析,确定主观权重,采用熵值法确定客观权重,并通过博弈论的组合赋权思想优化权重,充分挖掘权重信息,克服传统主、客观赋权方法的缺点.考虑到组合赋权的准确性,引入Kullback相对熵理论,验证了三者之间的一致性.同时利用属性识别理论对水质状况进行有效识别和比较分析,充分反映水质变化趋势.该模型应用于贵阳市饮用水源地水质评价,对比了三种不同评价方法.研究表明,该方法评价结果准确、说服性强,是一种实用、科学的评价方法.%Considering the defective tightness of relations among evaluation factors with traditional weighting method in comprehensive evaluation model for water quality,a concept of combination weighting was proposed.In the method,analytic hierarchy process was improved by introducing optimal transfer matrix concept to calculate subjective weights.Entropy method was used to calculate objective weights.Additionally,the combination weighting concept of game theory was used to optimize the weight,which excavated the full of weight information and conquered the drawback of traditional weighting method in subjective and objective weights.In order to ensure the accuracy of combination weights,Kullback relative entropy theory was introduced to verify the consistency among the three.At the same time,to fully reflect the variation trend of water quality at different points,water quality conditions were identificated effectively and analyzed comparatively by using attribute recognition theory.The model was applied to evaluate water quality of drinking water in Guiyang City.And three different evaluation methods were compared.The research demonstrated that the new method for water quality evaluation had accurate results and perfect performance,which was testified to be an effective and scientific one.