摘要
为弥补现有水质评价方法的不足之处,采用免疫进化算法优化投影寻踪模型的投影方向,提出基于免疫进化算法的投影寻踪模型,并将模型运用于我国五大湖水质评价及各种富营养化指标对湖泊富营养化程度评价的影响。结果表明通过比较基于免疫进化算法的投影寻踪水质评价模型的最佳投影方向各分量的大小和投影值的大小即可确定水体中污染物的主次及治理的优先并实现同类水之间的水质优劣比较;该模型能将高维数据通过最佳投影方向映射到低维子空间,便于对样本进行分类和综合评价,为水质评价研究领域提供了新的思路和方法。
Abstract.In order to remedy the shortcomings of present water quality assessment methods, the projection pursuit model based on the Immune Evolutionary Algorithm （IEA） is proposed to optimize projection direction. Then the model is used to evaluate the eutrophi- cation degree of five lakes in China. The results show that comparing the size of each component and the projection value of the model at the best projection direction can confirm the primary and secondary water pollutants and the management order and achieve water quality comparison between similar water. The model can map high dimensional data to the low-dimensional subspace through the best projection direction, which makes it is easy to conduct classification and comprehensive evaluation of the samples. It also pro- vides new ideas and methods for the research field of quality assessment.
出处
《节水灌溉》
北大核心
2012年第12期1-3,8,共4页
Water Saving Irrigation
基金
国家自然科学基金(D011103)
成都信息工程学院科研启动项目(KYTZ200919)
关键词
投影寻踪
免疫进化算法
水质评价
projection pursuit
Immune Evolutionary Algorithm
water quality assessment
