重庆邮电大学学报（自然科学版）2025，Vol.37Issue(1)：29-36,8.DOI:10.3979/j.issn.1673-825X.202401220018
吸收光谱法快速检测长江水COD的方法
Method for rapid detection of COD in Yangtze River water by absorption spectrometry
摘要
Abstract
To meet the high-precision detection requirements of chemical oxygen demand(COD)in surface water,this pa-per proposes a prediction model based on ultraviolet-visible absorption spectra combined with continuous projection-support vector machine(CP-SVM)regression.Using the absorption spectra of Yangtze River water as the research object,the SPXY method is used to divide the data into training and testing sets.Savitzky-Golay filtering is applied for data preprocess-ing,and the continuous projection algorithm is used to select characteristic wavelengths.Support vector machine regression is then used to fit the relationship between the selected characteristic wavelengths and sample concentrations,establishing a regression model for the chemical oxygen demand of Yangtze River water.The model's performance is evaluated using the coefficient of determination(R2),root mean square error(RMSE),and mean relative error,and compared with the pre-diction results of five other hybrid models.The model based on continuous projection-support vector machine regression shows the best performance,achieving rapid and accurate measurement of surface water COD.关键词
光谱学/吸收光谱/地表水/水质检测/化学需氧量Key words
spectroscopy/absorption spectrum/surface water/water quality testing/chemical oxygen demand分类
数理科学引用本文复制引用
郑培超,吕强,王金梅,曾金锐,李成林,杨琴..吸收光谱法快速检测长江水COD的方法[J].重庆邮电大学学报（自然科学版）,2025,37(1):29-36,8.基金项目
国家自然科学基金项目(32171627) （32171627）
重庆市教委科学技术研究项目重大项目(KJZD-M202200602)National Natural Science Foundation of China(32171627) （KJZD-M202200602）
Major Science and Technology Research Program of Chongqing Municipal Education Commission(KJZD-M202200602) （KJZD-M202200602）
