紫外-可见光谱水质多参数监测分析方法研究

Multi-parameter Monitoring Analysis Method Research of Water Quality Based on UV-Vis Spectroscopy

为了研究水质污染参数COD、TOC和NO_(3)-N的紫外-可见光谱吸收及交叉干扰特性,搭建了一套水质多参数监测系统,采用化学计量法和偏最小二乘回归算法建立解析模型,并对实际水样进行验证测试。实验表明:对浊度、COD、TOC、NO_(3)-N等溶液进行单因变量建模时,预测结果的R2优于0.9987;对浊度、COD和TOC混合溶液进行三因变量建模时,COD和TOC预测结果的R2分别为0.9576和0.9506;对浊度、COD、TOC和NO_(3)-N混合溶液进行四因变量建模时,NO_(3)-N预测结果的R2为0.9460;测量结果呈现良好的线性回归关系。

In order to study the UV-Vis spectral absorption and cross interference characteristics of turbidity,COD,TOC and NO_(3)-N,a multi-parameter monitoring system of water quality was set up,the chemical metering method and partial least squares regression algorithm were used to establish the corresponding analytical model,and the models were verified on the actual water samples.The results show that when using single dependent variable model to predict the concentration of the single solution such as turbidity,COD,TOC and NO_(3)-N,the R2 of the predicted results are better than 0.9987.When using three dependent variables model to predict the concentration of the mixed solutions of turbidity,COD and TOC,the R2 of the predicted results of COD and TOC are 0.9576 and 0.9506,respectively.When using four dependent variables model to predict the concentration of the mixed solutions of turbidity,COD,TOC and NO_(3)-N,the R2 of the predicted results of NO_(3)-N prediction is 0.9460.The results show a good linear regression relationship.