基于最小二乘支持向量机的水质COD参数的光谱解算方法研究  

Research on calculation method of COD parameter spectrum in water quality based on least squares support vector machine

作　　者：蔡鑫 吴绍锋[1] 李东波[1] Cai Xin;Wu Shaofeng;Li Dongbo(School of Mechanical and Electrical Engineering,Nanjing University of Science and Technology,Jiangsu Nanjing,210094,China)

机构地区：[1]南京理工大学机械工程学院

出　　处：《机械设计与制造工程》2019年第9期95-98,共4页Machine Design and Manufacturing Engineering

摘　　要：为研究利用光谱曲线解算水质COD(化学需氧量)参数的方法,首先利用COD标准液配置了一系列浓度的溶液,并采集了多组实际水样,然后根据朗伯-比尔定律分别测定其吸收光谱。使用回归型最小二乘支持向量机分别对COD标准液和实际水样进行建模,并分别使用两种模型对各自的测试样本进行预测比对。预测结果表明,最小二乘支持向量机在水质COD参数的光谱解算中具有较高的预测精度。In order to study the method of using spectral curve to calculate COD parameters in water quality,it prepares a series of COD standard liquids with different concentrations,collects several groups of actual water samples,measures their absorption spectra according to Lambert-Beer law.Using the regression least squares support vector machine,it builds the models of COD standard liquids and the actual water samples respectively,and predicts and compares the respective test samples.The prediction results show that the least squares support vector machine has good adaptability and high prediction accuracy in the COD parameter calculation of spectral water quality.

关 键 词：最小二乘支持向量机 吸光度 化学需氧量 

分 类 号：TH39[机械工程—机械制造及自动化]