Membrane pollution in-situ pollutant detection method based on front surface fluorescence data

Abstract

The invention relates to the technical field of membrane pollution detection, in particular to a membrane pollution in-situ pollutant detection method based on front surface fluorescence data, which adopts a machine learning model to carry out nonlinear modeling on original fluorescence data containing various interferences, and further introduces a symbol regression method. The prediction result of the machine learning model is analyzed, a membrane pollution in-situ pollutant relative concentration calculation formula is obtained, data correction is carried out on the relative concentration of the target pollutant, collection of multiple fluorescence parameters and correction of complex fluorescence data are not needed, and the accuracy of the detection result is improved. The method has the comprehensive advantages of high prediction precision, less parameter dependence, strong interference scene adaptation and the like, and is suitable for in-situ fluorescence analysis in the membrane pollution process.

Description

Technical Field

[0001] This invention relates to the field of membrane fouling detection technology, and in particular to an in-situ method for detecting contaminants in membranes based on front surface fluorescence data. Background Technology

[0002] Three-dimensional fluorescence analysis is characterized by its high efficiency, sensitivity, rich information content, and lack of the need for additional pretreatment, and has been widely used in recent years to characterize fluorescent natural organic compounds in water. Front-surface fluorescence technology involves setting the excitation beam at a 30°-60° incident angle with the sample surface to reduce interference from light absorption and scattering on fluorescence detection. During three-dimensional fluorescence detection, when the solution concentration meets the requirement of UV254 less than 0.05, the absorbance of the solution to the excitation light is very low, and the measured fluorescence intensity is directly proportional to the concentration of the fluorescent substance. As the solution concentration further increases, the linear relationship between fluorescence intensity and solution concentration gradually becomes invalid. When the solution concentration is too high, due to the internal filtration effect, the reabsorption of emitted light, and the influence of light scattering, a distortion occurs where the fluorescence intensity decreases with increasing concentration.

[0003] Due to interference factors such as internal filtering effects, emitted light reabsorption, and light scattering, the fluorescence data of the front surface requires further correction. Currently, there are three main categories of methods for correcting fluorescence data: one is physical model correction based on optical principles, such as the Kubelka-MonK theoretical correction model, which uses theoretical functions... Calculate the overall sustained-release coefficient, where The total mitigation coefficient, For the corresponding wavelength The diffuse reflection coefficient at that location is then adjusted using a correction function. Calculate the correction factor, where, For correction functions, The incident wavelength, To obtain the excitation wavelength, the original fluorescence data is divided by the correction coefficient to obtain the corrected fluorescence value. However, this type of method requires additional equipment to acquire reflectance data. In samples with high concentrations or complex compositions, using a single correction factor is often insufficient to accurately eliminate errors caused by internal filtration effects. When the type or concentration of absorbing substances in the sample changes, the correction factor needs to be re-measured and corrected. The second commonly used method is a linear regression model based on chemometrics or mathematics, such as principal component regression and partial least squares regression. This type of method has limited ability to model nonlinear systems. The correction set used for modeling must fully cover the variation range of all response components in the sample; otherwise, the model's generalization ability will be insufficient. The third commonly used method is a modeling and correction method based on machine learning. This method does not rely on physical optical models and does not require additional measurements of absorbance or reflectance. It has strong nonlinear modeling capabilities. However, traditional machine learning models are mostly "black box" models, whose internal decision-making processes are difficult to analyze and lack clear mathematical expressions, which is not conducive to mechanism research, knowledge transfer, and practical engineering application. Summary of the Invention

[0004] The purpose of this invention is to provide a method for in-situ detection of membrane contaminants based on front surface fluorescence data, thereby solving the aforementioned technical problems.

[0005] To achieve the above objectives, this invention provides a method for in-situ detection of membrane contaminants based on front surface fluorescence data, the specific steps of which are as follows: Step S1: Obtain the surface fluorescence dataset before membrane fouling and perform preprocessing; Step S2: Extract the feature input values ​​of the front surface fluorescence data. The feature input values ​​include the characteristic fluorescence intensity of the target pollutant and at least one Rayleigh scattering intensity related to the peak value of the target pollutant. Step S3: Construct a fluorescence feature dataset based on the extracted feature input values ​​and the known relative concentration values ​​of membrane fouling, and train at least one machine learning model using the fluorescence feature dataset and optimize the machine learning model. Select the optimal machine learning model using model evaluation metrics. Step S4: Construct a symbolic regression model based on the predicted output of the optimal machine learning model and the corresponding feature input values. Then, derive the formula for calculating the relative concentration of in-situ pollutants in membrane fouling based on the symbolic regression model. The formula for calculating the relative concentration of in-situ pollutants in membrane fouling is as follows:

[0006] Where Y represents the relative concentration of in-situ pollutants in membrane fouling. To correct the parameters, To detect the fluorescence intensity of the target pollutant, The scattering intensity at the excitation wavelength of the target pollutant; Step S5: Collect surface fluorescence data before membrane fouling, and calculate the relative concentration of in-situ pollutants in membrane fouling according to the formula for calculating the relative concentration of in-situ pollutants in membrane fouling.

[0007] Preferably, in step S1, the process of obtaining the surface fluorescence dataset before membrane fouling is as follows: Step S11: Prepare target pollutant solutions with different relative concentrations and add different types of interfering substances to the target pollutant solutions; Step S12: Use an ultrafiltration membrane to filter a solution of target pollutants containing different interfering substances, and collect three-dimensional fluorescence spectral data of the front surface of the ultrafiltration membrane using a fiber optic probe in a light-protected environment; Step S13: Perform data preprocessing on the three-dimensional fluorescence spectroscopy data. Data preprocessing includes outlier removal, standardization, or normalization.

[0008] Preferably, the target pollutant solution is algal solution, and the interfering substances are one or any combination of nano-sized silica, humic acid, aluminum chloride, ferric sulfate, organic pigment red, organic pigment green, and organic pigment purple.

[0009] Preferably, the fluorescence spectroscopy detection range is 610nm-660nm for excitation wavelength and 610nm-660nm for emission wavelength, the excitation light data interval is 2-10nm, the emission light data interval is 0.5-5nm, the excitation light slit width is 5-20nm, the emission light slit width is 5-20nm, and the scanning speed is 3000-60000nm / min.

[0010] Preferably, the ultrafiltration membrane is a polyvinylidene fluoride flat sheet ultrafiltration membrane.

[0011] Preferably, in step S3, the machine learning model is one or any combination of an extreme gradient boosting tree model, a random forest model, a stacked ensemble model, a multilayer perceptron-artificial neural network model, a K-nearest neighbor model, or a support vector regression model.

[0012] Preferably, in step S3, the optimization of the machine learning model includes: Optimize the number of trees, tree depth, learning rate, sample sampling ratio, feature sampling ratio, minimum split gain, and regularization parameters for extreme gradient boosting tree models; The stacked ensemble model is cross-validated and integrated using a meta-model. Optimize the number of layers, number of neurons, learning rate, and regularization parameters for the multilayer perceptron-artificial neural network model; The K-nearest neighbor model optimizes the K value, distance metric, and weighted voting parameters. The kernel function, regularization coefficient, and tolerance range parameters of the support vector regression model are optimized by combining grid search and cross-validation for parameter tuning.

[0013] Preferably, the model evaluation metrics include goodness of fit, mean squared deviation, mean absolute deviation, root mean square error, and mean relative error.

[0014] Preferably, the formula for calculating the relative concentration of in-situ pollutants in membrane fouling is selected from the output of the symbolic regression model based on the algal absorption interference compensation mechanism and fitting accuracy.

[0015] Therefore, the present invention employs the above-mentioned in-situ contaminant detection method for membrane fouling based on front surface fluorescence data, which has the following beneficial effects: (1) By introducing a machine learning model, the complex nonlinear relationship between the characteristic fluorescence peaks of the target pollutant in the original three-dimensional fluorescence signal and interferences such as Rayleigh scattering in its neighboring region is modeled to obtain high-precision relative concentration prediction results. The high-precision relative concentration prediction results are introduced into the symbolic regression model to explore the analytical function relationship between the input features and the output results. No additional measurement of optical parameters such as absorbance and reflectance is required, and no complex correction of fluorescence data is required. The data correction of in-situ pollutants in membrane fouling can be completed solely based on the original three-dimensional fluorescence data.

[0016] (2) The characteristic fluorescence intensity of the specific target pollutant and the Rayleigh scattering signal in its neighborhood are combined as key feature inputs, and different kinds of interference are added to the training dataset to enhance the model’s ability to extract effective information under complex interference background, thereby efficiently completing the algae density modeling and interference correction tasks, and improving the model’s robustness, generalization ability and actual prediction accuracy.

[0017] The technical solution of the present invention will be further described in detail below with reference to the accompanying drawings and embodiments. Attached Figure Description

[0018] Figure 1 This is a flowchart of an in-situ contaminant detection method for membrane fouling based on front surface fluorescence data according to the present invention. Figure 2 A histogram showing the goodness of fit for the first group, which uses pure algae solution without any added interfering substances. Figure 3 A histogram showing the goodness of fit for the algal solution with added interfering substances in the second group; Detailed Implementation

[0019] In the description of this invention, it should be noted that the terms "upper," "lower," "inner," and "outer," etc., indicate the orientation or positional relationship based on the orientation or positional relationship shown in the accompanying drawings, or the orientation or positional relationship commonly used when the product is in use. They are used only for the convenience of describing the invention and simplifying the description, and do not indicate or imply that the device or element referred to must have a specific orientation, or be constructed and operated in a specific orientation. Therefore, they should not be construed as limitations on the invention. In the description of this invention, it should also be noted that, unless otherwise explicitly specified and limited, the terms "set," "install," and "connect" should be interpreted broadly. For example, they can refer to a fixed connection, a detachable connection, or an integral connection; they can refer to a mechanical connection or an electrical connection; they can refer to a direct connection or an indirect connection through an intermediate medium; and they can refer to the internal communication between two components. Those skilled in the art can understand the specific meaning of the above terms in this invention based on the specific circumstances.

[0020] The embodiments of the present invention will now be described in detail with reference to the accompanying drawings.

[0021] like Figure 1 As shown, a method for in-situ detection of membrane contaminants based on front surface fluorescence data is described, with the following specific steps: Step S1: Obtain the surface fluorescence dataset before membrane fouling and perform preprocessing.

[0022] In step S1, the process of obtaining the surface fluorescence dataset before membrane fouling is as follows: Step S11: Prepare target pollutant solutions with different relative concentrations and add different types of interfering substances to the target pollutant solutions. In this embodiment, the target pollutant solution is an algal solution, which is prepared using raw water from Microcystis aeruginosa from the freshwater algae bank of the Institute of Hydrobiology, Chinese Academy of Sciences. The interfering substances are one or any combination of nano-sized silica, humic acid, aluminum chloride, ferric sulfate, organic pigment red, organic pigment green, and organic pigment purple.

[0023] Step S12: Filter the target pollutant solution containing different interfering substances using an ultrafiltration membrane. The ultrafiltration membrane is a polyvinylidene fluoride (PVDF) flat sheet ultrafiltration membrane, manufactured by Guochu Technology (Xiamen) Co., Ltd., with an effective filtration area of ​​45 cm². 2 After filtering a fixed volume of algal solution, three-dimensional fluorescence spectral data of the front surface of the ultrafiltration membrane were collected using an optical fiber probe under light-protected conditions. The fluorescence spectral detection range was 610nm-660nm excitation wavelength, 610nm-660nm emission wavelength, with a scanning interval of 5nm excitation wavelength, 2nm emission wavelength, 10nm slit width, and a scanning speed of 30000nm / s.

[0024] Step S13: Perform data preprocessing on the three-dimensional fluorescence spectroscopy data. Data preprocessing includes outlier removal, standardization or normalization, converting the exported txt file of the three-dimensional fluorescence spectroscopy data into an easily readable and recognizable excel format, and performing relevant preprocessing according to the actual needs of subsequent models.

[0025] Step S2: Extract the feature input values ​​from the front surface fluorescence data. The feature input values ​​include the characteristic fluorescence intensity of the target pollutant and the Rayleigh scattering intensity correlated with the peak values ​​of the two target pollutants. The Rayleigh scattering intensity correlated with the peak values ​​of the two target pollutants is obtained by extracting the Rayleigh scattering fluorescence intensity at excitation wavelength / emission wavelength of 610 nm / 610 nm and the characteristic fluorescence intensity of *Microcystis aeruginosa* at 610 nm / 660 nm from the fluorescence spectral data as feature input values.

[0026] The true algal cell density was determined by measuring the absorbance of the pure algal solution using a UV-Vis spectrophotometer at a wavelength of 685 nm. Based on the existing formula relating *Microcystis aeruginosa* cell concentration to absorbance, the cell concentration was calculated, and then the cell density was calculated based on the membrane area. The solution for determining the true algal cell density involves first measuring the absorbance of the pure algal solution using a UV-Vis spectrophotometer at a wavelength of 685 nm, and then applying the existing formula relating *Microcystis aeruginosa* cell concentration to absorbance. The cell concentration of the algal solution is calculated, and then the algal cell density is calculated based on the membrane area.

[0027] The input and output values ​​of the dataset are labeled and classified. Among the input values, the Rayleigh scattering fluorescence intensity at excitation wavelength / emission wavelength of 610nm / 610nm is x1, the characteristic fluorescence intensity of Microcystis aeruginosa at excitation wavelength / emission wavelength of 610nm / 660nm is x2, and the Rayleigh scattering fluorescence intensity at excitation wavelength / emission wavelength of 660nm / 660nm is x3. The true algal cell density is y.

[0028] Step S3: Construct a fluorescence feature dataset based on the extracted feature input values ​​and the known relative concentration of membrane fouling. Divide the fluorescence feature dataset into a training set and a test set in an 8:2 ratio. Train and test six machine learning models using the fluorescence feature dataset, optimize the machine learning models, and select the optimal machine learning model using model evaluation metrics.

[0029] Six machine learning models were used: an extreme gradient boosting tree model, a random forest model, a stacked ensemble model, a multilayer perceptron-artificial neural network model, a K-nearest neighbor model, and a support vector regression model. For the extreme gradient boosting tree model, the number of trees, tree depth, learning rate, sample sampling ratio, feature sampling ratio, minimum split gain, and regularization parameters were optimized. For the stacked ensemble model, cross-validation was performed using a meta-model. For the multilayer perceptron-artificial neural network model, the number of layers, number of neurons, learning rate, and regularization parameters were optimized. For the K-nearest neighbor model, the K value, distance metric, and weighted voting parameters were optimized. For the support vector regression model, the kernel function, regularization coefficient, and tolerance range parameters were optimized, and parameter tuning was performed using grid search and cross-validation to control generalization error.

[0030] Model evaluation metrics include goodness of fit R. 2 Mean square deviation (MSE), mean absolute deviation (MAE), root mean square error (RMSE), and mean relative error (MeanRE).

[0031] Step S4: Based on the prediction results dataset of the optimal machine learning model and the corresponding feature input values, construct a symbolic regression model to obtain the formula for calculating the relative concentration of in-situ pollutants in membrane fouling. Use the SymbolicRegression.jl symbolic regression tool to model and analyze the data exported above. Select the formula for calculating the relative concentration of in-situ pollutants in membrane fouling from the output of the symbolic regression model based on the algal absorption interference compensation mechanism and fitting accuracy.

[0032] The formula for calculating the relative concentration of in-situ pollutants in membrane fouling is as follows:

[0033] Where Y represents the relative concentration of in-situ pollutants in membrane fouling. To correct the parameters, To detect the fluorescence intensity of the target pollutant, The scattering intensity at the excitation wavelength of the target pollutant; Step S5: Collect surface fluorescence data before membrane fouling, and calculate the relative concentration of in-situ pollutants in membrane fouling according to the formula for calculating the relative concentration of in-situ pollutants in membrane fouling.

[0034] A comparative experiment was conducted to verify the anti-interference capability of the technical solution in this embodiment: The first group used pure algal solution without any interfering substances, while the second group used algal solution with interfering substances. The methods employed included no correction, a Kubelka-MonK theoretical correction model, and the technique described in this embodiment. The Kubelka-MonK theoretical correction model first considers x1 and x3 as... , Then through theoretical functions Calculate the overall sustained-release coefficient, and then use the correction function. The correction coefficient is calculated, and the original fluorescence data is divided by the correction coefficient to obtain the fluorescence value after KM theoretical correction, which is then fitted to the algal cell density.

[0035] Goodness of fit, such as Figures 2-3 As shown, in pure algae solution, the goodness of fit of the three techniques is similar, but the performance increases in that order. In the presence of interfering substances (a combination of nano-sized silica, humic acid, aluminum chloride, ferric sulfate, organic pigment red, organic pigment green, and organic pigment purple), the goodness of fit of the model without correction and using the Kubelka-MonK theory to correct the model decreases to 0.507 and 0.534, respectively, while the goodness of fit of the technical solution in this embodiment is 0.8847, which shows high reliability and strong adaptability to interference scenarios.

[0036] Finally, it should be noted that the above embodiments are only used to illustrate the technical solutions of the present invention and not to limit them. Although the present invention has been described in detail with reference to preferred embodiments, those skilled in the art should understand that modifications or equivalent substitutions can still be made to the technical solutions of the present invention, and these modifications or equivalent substitutions cannot cause the modified technical solutions to deviate from the spirit and scope of the technical solutions of the present invention.

Claims

1. A method for in-situ detection of membrane contaminants based on front surface fluorescence data, characterized in that, The specific steps are as follows: Step S1: Obtain the surface fluorescence dataset before membrane fouling and perform preprocessing; Step S2: Extract the feature input values ​​of the front surface fluorescence data. The feature input values ​​include the characteristic fluorescence intensity of the target pollutant and at least one Rayleigh scattering intensity related to the peak value of the target pollutant. Step S3: Construct a fluorescence feature dataset based on the extracted feature input values ​​and the known relative concentration values ​​of membrane fouling, and train at least one machine learning model using the fluorescence feature dataset and optimize the machine learning model. Select the optimal machine learning model using model evaluation metrics. Step S4: Construct a symbolic regression model based on the predicted output of the optimal machine learning model and the corresponding feature input values. Then, derive the formula for calculating the relative concentration of in-situ pollutants in membrane fouling based on the symbolic regression model. The formula for calculating the relative concentration of in-situ pollutants in membrane fouling is as follows: Where Y represents the relative concentration of in-situ pollutants in membrane fouling. To correct the parameters, To detect the fluorescence intensity of the target pollutant, The scattering intensity at the excitation wavelength of the target pollutant; Step S5: Collect surface fluorescence data before membrane fouling, and calculate the relative concentration of in-situ pollutants in membrane fouling according to the formula for calculating the relative concentration of in-situ pollutants in membrane fouling.

2. The method for in-situ detection of membrane contaminants based on front surface fluorescence data according to claim 1, characterized in that: In step S1, the process of obtaining the surface fluorescence dataset before membrane fouling is as follows: Step S11: Prepare target pollutant solutions with different relative concentrations and add different types of interfering substances to the target pollutant solutions; Step S12: Use an ultrafiltration membrane to filter a solution of target pollutants containing different interfering substances, and collect three-dimensional fluorescence spectral data of the front surface of the ultrafiltration membrane using a fiber optic probe in a light-protected environment; Step S13: Perform data preprocessing on the three-dimensional fluorescence spectroscopy data. Data preprocessing includes outlier removal, standardization, or normalization.

3. The method for in-situ detection of membrane contaminants based on front surface fluorescence data according to claim 2, characterized in that: The target pollutant solution is algal solution, and the interfering substances are one or any combination of nano-sized silica, humic acid, aluminum chloride, ferric sulfate, organic pigment red, organic pigment green, and organic pigment purple.

4. The method for in-situ detection of membrane contaminants based on front surface fluorescence data according to claim 3, characterized in that: The fluorescence spectroscopy detection range is 610nm-660nm for excitation wavelength and 610nm-660nm for emission wavelength. The excitation data interval is 2-10nm, the emission data interval is 0.5-5nm, the excitation slit width is 5-20nm, the emission slit width is 5-20nm, and the scanning speed is 3000-60000nm / min.

5. The method for in-situ detection of membrane contaminants based on front surface fluorescence data according to claim 4, characterized in that: The ultrafiltration membrane is a polyvinylidene fluoride flat sheet ultrafiltration membrane.

6. The method for in-situ detection of membrane contaminants based on front surface fluorescence data according to claim 5, characterized in that: In step S3, the machine learning model is one or any combination of extreme gradient boosting tree model, random forest model, stacked ensemble model, multilayer perceptron-artificial neural network model, K-nearest neighbor model or support vector regression model.

7. The in-situ contaminant detection method for membrane fouling based on front surface fluorescence data according to claim 6, characterized in that: In step S3, the optimization of the machine learning model includes: Optimize the number of trees, tree depth, learning rate, sample sampling ratio, feature sampling ratio, minimum split gain, and regularization parameters for extreme gradient boosting tree models; The stacked ensemble model is cross-validated and integrated using a meta-model. Optimize the number of layers, number of neurons, learning rate, and regularization parameters for the multilayer perceptron-artificial neural network model; The K-nearest neighbor model optimizes the K value, distance metric, and weighted voting parameters. The kernel function, regularization coefficient, and tolerance range parameters of the support vector regression model are optimized by combining grid search and cross-validation for parameter tuning.

8. The method for in-situ detection of membrane contaminants based on front surface fluorescence data according to claim 7, characterized in that: Model evaluation metrics include goodness of fit, mean squared deviation, mean absolute deviation, root mean square error, and mean relative error.

9. The in-situ contaminant detection method for membrane fouling based on front surface fluorescence data according to claim 8, characterized in that: The formula for calculating the relative concentration of in-situ pollutants in membrane fouling was selected from the output of the symbolic regression model based on the algal absorption interference compensation mechanism and fitting accuracy.

Images