Font Size: a A A

Study On The Prediction Model Of Nitrate Nitrogen And Turbidity Compensation In Water Based On Ultraviolet Spectrum

Posted on:2020-05-22Degree:MasterType:Thesis
Country:ChinaCandidate:L HeFull Text:PDF
GTID:2381330599960440Subject:Instrumentation engineering
Abstract/Summary:PDF Full Text Request
In recent years,water pollution has become more and more serious.Nitrate nitrogen pollution is one of the main reasons.Excessive nitrate nitrogen content in water can cause eutrophication of water body and damage the ecological environment.When the content of nitrate nitrogen in drinking water is too high,nitrate nitrogen will be reduced to nitrite,which poses a great threat to human health.Therefore,the detection of this pollutant is pretty important.Ultraviolet absorption spectroscopy can be used to detect nitrate nitrogen,and has the advantages of fast detection speed and no need of chemical reagents to cause secondary pollution.Ultraviolet absorption spectra of nitrate nitrogen,turbidity and their mixed solutions were collected by the ultraviolet spectrophotometer in the laboratory,and their spectral analysis was carried out.The main research work is as follows:Firstly,a quantitative analysis model based on the combination of Interval Partal Least Squares(iPLS)and Partial Least Squares Support Vector Machine(LS-SVM)is proposed.Through the spectral integral analysis of all samples,the critical concentration value was screened out,and the samples were divided into low concentration and high concentration to establish different prediction models.A linear regression prediction model of iPLS was established for low-concentration samples with good spectral linearity.A prediction model of ls-svm was established for high-concentration samples with severe spectral overlap.Then,the model fusion between iPLS and LS-SVM is carried out and compared with other quantitative analysis models.The square(R~2)of the correlation coefficient between the predicted value and the real concentration is 0.9998,the root mean square error(RMSEP)is0.428,and the mean absolute error(MAE)is 4.27%.The prediction effect is obviously better than other quantitative analysis models.Secondly,Considering that the ultraviolet absorption spectra are easily disturbed by turbidity,the ultraviolet absorption spectra of nitrate nitrogen and mixed solutions with different turbidity are analyzed experimentally.By analyzing the normalized spectra of the mixed solution,the normalized spectra of the bands of 250~260nm and335~370nm basically overlapped,and the regression model of spectral integral and turbidity value of the two bands was established.The turbidity value in the mixed solution can be calculated by this regression model,and then the turbidity compensation can be carried out accordingly.Finally,a method of nitrate nitrogen turbidity compensation based on compensation curve is proposed.The effect of turbidity on the absorption spectrum of nitrate nitrogen was characterized by introducing a compensation coefficient k _N(?)between 0 and 1.The compensation coefficients of different turbidity in the main absorption spectral region of nitrate nitrogen are calculated based on the spectral data collected from experiments,and the compensation curves of different turbidity can be obtained.After the turbidity value in mixed solution is obtained by turbidity calculation model,the turbidity compensation curve of the turbidity value is used to compensate the turbidity of nitrate nitrogen absorption spectrum.Compared with other compensation methods,the compensation effect is obviously superior to other methods.
Keywords/Search Tags:UV absorption spectrum, nitrate nitrogen, hybrid prediction model, turbidity compensation
PDF Full Text Request
Related items