模糊识别在水质综合评价中的应用

Application of fuzzy recognition in comprehensive evaluation of water quality

应用模糊数学方法对水质进行综合评价,其关键问题是如何合理地确定各评价指标的权重.将最优权法引入模糊评价以确定各评价指标的权系数,通过求解权向量的最优解来确定各指标权重,可避免确定评价指标权重的主观任意性,增加评价结果的分辨性.实例计算与结果对比表明,该方法克服了各指标的权系数随实测样本的不同而变化和评价结果分辨性差的缺点,提高了模型的通用性和评价结果的可比性,同时也简化了多次重复确定权系数的工作量.

The method of fuzzy mathematics used in comprehensive evaluation of water quality is objective, but the key problem is how to quantify the weights of different evaluation indexes in fuzzy system reasonably. A new approach of optimizing weight is proposed to determine the weight of evaluation indexes by fuzzy recognition. This method quantifies the weights of all indexes through seeking for the optimal solution of the weight vector, which can avoid the subjective randomicity and increase the distinguishability of results. The cases and analysis results show that this method overcomes some disadvantages, such as the change of indexes weights with the variety of different samples which goes against the compare to the comprehensive evaluation results, and the classification of evaluation results is unconspicuous. In addition, this method increases the generalization of the model, enhances the comparability of evaluated results, as well as to decrease the work of multi-computing to determine the weight.