紫外光谱结合间隔偏最小二乘法测定短程硝化过程中硝酸盐氮和亚硝酸盐氮的含量

Shortcut Nitrification Ultraviolet Spectrum Analysis in SBRProcess of Fast Start

采用序批式活性污泥反应器(SBR)研究了短程硝化反应系统启动过程中,紫外光谱预测反应器出水中硝酸盐氮和亚硝酸盐氮的含量可行性。利用实验室配制不同浓度梯度的硝酸盐氮和亚硝酸盐氮混合溶液,以及反应器出水样品,采用间隔偏最小二乘法建立了样品中紫外光谱与硝酸盐氮和亚硝酸盐氮含量的定量分析模型。结果表明:采用实验室配制硝酸盐氮和亚硝酸盐氮混合溶液时校正集相关系数分别达到0.9993、0.9283,校正均方根误差分别为0.0257、0.3596,预测值与真实值的误差范围为-6.672%~9.097%和-21.839%~17.645%。采用反应器出水时校正集相关系数分别达到0.9799、0.9750,校正均方根误差分别为2.8847、4.5015,预测值与真实值的误差范围为-4.445%~5.505%和-7.791%~14.484%。模型总体预测效果较好,说明紫外光谱可以实现对生物短程硝化出水样品进行快速定性和定量分析。

Using sequencing batch type activated sludge reactor （SBR）, this paper studied shortcut nitrification reaction system startup process and the ultraviolet spectrum predicted reactor out the content of nitrate nitrogen and nitrite nitrogen in water. Using laboratory from different concentration gradient of nitrate nitrogen and nitrite nitrogen mixture and the reactor effluent samples, interval partial least squares method is adopted to establish the ultraviolet spectrum and nitrate nitrogen and nitrite nitrogen content of quantitative analysis model.The results showed that the calibrating correlation coefficientof nitrate nitrogen and nitrite nitrogen mixture prepared by the laboratoryis 0.9993 and 0.9283, respectively;root mean square error correction were 0.0257, 0.3596; error range of the predicted values and the real value is 6.672 - 9.097 % and 21.839 - 17.645 %. Using the reactor effluent calibratingcorrelation coefficient of 0.9799, 0.9750 respectively, root mean square error correction were 2.8847, 4.5015 ; error range of the predicted values and the real value was 4.445 - 5.505% and 7.791 - 14.484%. The overall prediction effect of the model is better, indicating that ultraviolet spectrum can be realized on shortcut nitrification water samples for rapid qualitative and quantitative analysis.