摘要
水体污染威胁着人类赖以生存的环境。污染物类型的快速分类识别能为水质预警及应急处理提供决策支持。通过测量水质参数:电导率(K)、温度(T)、酸碱度(pH)、氧化还原电位(ORP)、溶解氧(DO)和紫外吸光度(UV_(254)),以余弦距离比值定义的"区分度"作为相似性测度对3种类型的水体污染物进行分类。结果表明,采用区分度对污染物进行分类,正确率达95.9%,比采用余弦距离方法高近3%。该方法不仅增大了不同类别污染物之间的差异,而且具有耗时短、操作简便、准确率高等优点。
Water pollution has a significant impact on human living environment. Quick classification of the contaminants is very important to provide support for water quality early warning and remediation attempts. By measurement of water quality parameters, such as conductivity（K）, temperature（T）, acidity（pH）, oxidation reduction potential（ORP）, dissolved oxygen（DO） and UV absorbance at 254 nm（UV（254））, a new method of water contaminants classification has been proposed. Three kinds of contaminants were determined by using the cosine discrimination. The results show the class of these contaminants can be well identified with 95.9% accuracy by proposed method, which is almost 3% higher than the cosine distance method.The proposed method not only increases the classification accuracy in contaminant discrimination, but has the advantages in shortening measure time and simplifying experimental operation as well.
出处
《环境科学与技术》
CAS
CSCD
北大核心
2016年第10期99-105,共7页
Environmental Science & Technology
基金
科技部国家重大科学仪器设备开发专项基金项目(2011YQ150078)
关键词
水质参数
区分度
余弦距离
污染物分类
water quality parameters
discrimination
cosine distance
contaminant classification
