主成分分析用于管网水质综合评价的研究

Study on Application of Principal Component Analysis in Comprehensive Evaluation of Water Quality of Network

为明确供水管网中各种水质指标的关系,并能够综合各种水质监测指标来评价管网水的总体质量,利用基于多元统计分析原理的主成分分析法,将多种水质监测指标纳入同一系统进行定量化研究。以北方A市某小区供水管网为例,对其水质监测数据进行主成分分析,初步了解了该小区管网水质的变化情况,并按照总体水质情况对管网水质监测点进行了排序。此外,还按照水质指标间相关性的大小将所选定的13项常规监测指标分成了3组,为建立相应的宏观模型奠定了基础。

In order to clarify the correlation among different water quality indexes in water supply network and synthesize different indexes for monitoring the water quality to evaluate the total water quality of network, a quantitative study was conducted using principal component analysis method based on mult- ivariate statistic analysis principle and incorporating various indexes for monitoring the water quality into the same system. Taking the water supply network of a community in northern city A, China as the exam- ple, the monitored water quality data were processed using the principal component analysis method. After knowing the basic water quality changes in this community＇ s network, the monitoring spots were listed according to their water quality levels. According to the correlation degree, the chosen thirteen conventional monitoring indexes were divided into three groups, thus laying a foundation for establishing the relevant macroscopic model.