基于粒子群算法湖泊富营养化评价的投影寻踪方法  

Comprehensive Evaluation Method of Projection Pursuit Based on Particle Swarm Optimization in Lake Eutrophication

作　　者：张春乐[1] 方崇[1] 黄伟军[1] 

机构地区：[1]广西水利电力职业技术学院,广西南宁530023

出　　处：《安徽农业科学》2010年第27期14823-14825,14830,共4页Journal of Anhui Agricultural Sciences

基　　金：国家自然科学基金项目(No.50679047);广西水利水电科技计划项目(桂水科合字2005-2号)

摘　　要：根据湖泊富营养化特性,选择总磷(TP)、总氮(TN)、耗氧量(CODMn)、透明度(SD)、生物量等5个评价指标作为评价因子,建立湖泊富营养化评价的投影寻踪分析模型,采用粒子群算法对评价模型进行优化,并将该模型应用于我国8个湖泊富营养化程度的评价。研究表明,投影寻踪回归分析法,避免了传统评价方法由于主观原因造成的误差,评价结果合理可信、方法简单。Based on characteristics of lake eutrophication,total phosphorus（TP）,total nitrogen（TN）,oxygen consumption（CODMn）,transparency（SD）and biomass were chosen as evaluation factors.Projection pursuit analysis model of evaluation on lake eutrophication was established.The evaluation model was optimized using particle swarm optimization,and the model was applied to analyze eight lakes.The evaluation results were accurate.This method is simple,and some errors are eliminated because the subjective factors of traditional methods are avoided.Projection pursuit is a new evaluation method of lake eutrophication,and it shows a wide application prospect.

关 键 词：投影寻踪 粒子群算法 湖泊 富营养化 综合评价 

分 类 号：X524[环境科学与工程—环境工程]