摘要
给出了一种空间向量遗传聚类分析方法,对海洋环境监测得到的多参数数据进行分析。采用空间向量遗传聚类方法对采样点的温度,盐度,pH,DO等参数进行聚类,并将聚类结果投影到环境监测参数特征空间,便可以在特征空间中直观地对监测区某一时段采样点进行多参数数据分析,获知各采样点水质状况。通过对不同时段采样点数据的聚类分析,还可以对监测区海水变化趋势进行判断。此方法不仅能挖掘出采样点数据的关联性,而且使得对多采样点多参数数据的分析变得直观、清晰,提高了对海洋环境监测数据的分析效果。
A method based on genetic clustering of space vector is proposed for analyzing the multi-parameters data in marine environmental monitoring.The multi-parameters data of sampling points including temperature,salinity,pH and DO is clustered by this method,with the clustering results projected into feature space of environment monitoring parameters,the data analysis in the monitoring area can be intuitively analyzed then.The water quality condition can be known by analysis of data at certain time.The variable trend of seawater in monitoring area can be estimated,with analysis of clustering results for the sample data in different periods of time.By this way,not only the relevance between sampled data can be mined,but also analysis results can be displayed visually and clearly.
出处
《传感器与微系统》
CSCD
北大核心
2012年第6期22-24,28,共4页
Transducer and Microsystem Technologies
基金
上海市科委科研计划资助项目(10510502800)
关键词
遗传聚类
多参数
特征空间
海洋环境监测
genetic clustering
multi-parameters
feature space
marine environmental monitoring