基于粒子群算法的混合尘溯源解析技术改进  

Improvement of source apportionment by exploring origin of mixed dust source with particle swarm optimization

作　　者：易柯欣 邹长武[1] 刘伟[1] 王皓[1] 李友平[1,2] 

机构地区：[1]成都信息工程大学资源环境学院,四川成都610225 [2]西华师范大学环境科学与工程学院,四川南充637009

出　　处：《中国环境科学》2015年第11期3247-3250,共4页China Environmental Science

基　　金：国家自然科学基金(41405036);高原大气与环境四川省重点实验室开放课题(PAEKL-2013-Y4);化学合成与污染控制四川省重点实验室项目(CSPC2014-4-2)

摘　　要：采用粒子群算法代替智能解域搜索算法进行CMB模型优化求解,提出改进混合尘溯源解析技术,并结合实例对改进混合尘溯源解析技术的解析结果与混合尘溯源解析技术进行了比较.结果显示,改进混合尘溯源解析技术解析得到的扬尘贡献率为28.01%,低于混合尘溯源解析技术的28.75%,计算得到的受体成分谱中各元素的计算值/实测值较混合尘溯源解析技术更接近1,表明改进混合尘溯源解析技术的解析结果更加准确、合理.Improved exploring origin of mixed dust source（IEOMDS） was proposed by using particle swarm optimization（PSO） to calculate the contributions of all dust instead of search solution space with intelligence（SSSI）, which was first applied in the method of exploring origin of mixed dust source（EOMDS） in CMB model. After that, IEOMDS was tested in source apportionment of atmospheric particulates for a city, and compared its results with EOMDS. The results showed that the contribution rate of dust according to IEOMDS model was 28.01%, which was lower than 28.75% according to the original model, and the ratios of calculated data and measured data of receptor elements based on IEOMDS model were closer to 1, indicating that the results of IEOMDS model are more accurate and reasonable.

关 键 词：混合尘溯源解析技术 智能解域搜索算法 粒子群算法 改进混合尘溯源解析技术 

分 类 号：X513[环境科学与工程—环境工程]