投影寻踪模型用于水质等级评价的研究  

Study on Projection Pursuit Model Applied in Water Quality Assessment

作　　者：刘建[1] 刘丹[1] 

机构地区：[1]西南交通大学环境科学与工程学院,四川成都610031

出　　处：《安徽农业科学》2009年第1期261-262,265,共3页Journal of Anhui Agricultural Sciences

基　　金：四川省交通厅科学技术研究项目(05009)

摘　　要：投影寻踪模型具有可处理高维、非正态数据分类问题的优点,适合于在水质等级评价中应用。该文在介绍该模型基本计算步骤的基础上,经建模分析,得出:各评价等级取值范围内样本个数不小于20即可具有较好的稳定性,投影向量分量的取值范围为[0,1]时更有利于提高计算速度,选用加速遗传算法寻优时间更短,使用图解法或直接比较法更为方便和快捷。这些探索能为该模型在今后的实际应用提供有益参考。Projection pursuit model（PP）,being good at dealing with high dimension and non-normal data,was suitable for water quality assessment.This paper built water quality assessment model based on some basic introduction to PP.It made several conclusions as follows： PP model was stable when sample size was no less than 20.It run much faster when the range of projection vector was constrained between 0 and 1.Accelerated genetic algorithm should be recommended because of its unique advantages.Using scatter diagram of making comparison directly was more convenient than using regression model.All these reflections were expected to provide some useful references for the use of PP in the future.

关 键 词：投影寻踪 水质评价 遗传算法 样本数量 

分 类 号：X52[环境科学与工程—环境工程]