KR101430800B1 - Method for prediction and diagnosis of membrane fouling using combined-mechanistic fouling model and system thereof - Google Patents

Abstract

The present invention is to grasp the effect of a plurality of mechanisms integrally on actual membrane contamination, to create a new membrane contamination integrated model for predicting membrane contamination phenomenon, and to integrate four membrane contamination models into one expression And to provide a method and system for predicting and diagnosing membrane fouling using an integrated model.
The membrane fouling predicting method using the integrated model according to the present invention is a method for predicting membrane fouling by using an integrated modeling unit to calculate the transmembrane pressure differential (TMP) using a full pore blocking model, a medium pore blocking model, a cake formation model and a standard pore blocking model ; The data collection unit periodically collecting actual data of membrane contamination including inter-membrane pressure differential data; Calculating the inter-membrane pressure difference of the integrated model; Calculating an optimization parameter for the integrated model using the inter-membrane pressure difference of the membrane-contaminated actual data and the inter-membrane pressure difference of the integrated model; And the film fouling predicting section calculates the inter-membrane pressure difference prediction value of the membrane contamination integrated model using the optimization parameter.

Description

[0001] METHOD FOR PREDICTION AND DIAGNOSIS OF MEMBRANE FOULING USING COMBINED-MECHANISTIC FOULING MODEL AND SYSTEM THEREOF [0002]

The present invention relates to a method and a system for generating a membrane contamination integrated model by integrating a single model describing membrane contamination phenomenon, and using this to predict and diagnose membrane contamination phenomenon.

A membrane separates a substance to be treated contained in a solution by a liquid or solid membrane or a pore diameter (several tens of nanometers to several tens of micrometers) capable of separating the mixture by selectively passing a specific component and surface particles It is a membrane.

Membrane contamination phenomenon occurs in which the permeation flux decreases sharply as the filtration time increases when treating the wastewater using the membrane. This is because the suspended solids contained in the wastewater or the substances easily adsorbed on the surface of the separation membrane are accumulated on the membrane surface and pores to reduce the flow. Such membrane fouling is caused by the interaction between the material to be separated and the membrane material, and the size of the permeate and the size of the membrane pore are closely related. In addition, membrane contamination is affected by membrane type and operating conditions. Combustion occurs due to various complex factors. This phenomenon results in a decrease in permeation flux and an increase in rejection rate.

1 is a schematic diagram of a static pressure plugging filtration model showing a decrease in permeation flux with a change in filtration time. The static pressure plugging filtration model is derived from the assumption that the pore size is uniform, and there are pore blocking models and cake formation models.

In the complete pore blocking model, the solute particle size and the pore size of the membrane are the same, so that the solute particle completely blocks the pores, but the number of pores decreases in the form that the particles do not overlap each other, It is proportional to volume.

Intermediate pore blocking model is an intermediate form of complete shielding model and cake formation model. Particles reaching membrane surface do not necessarily block each pore, but they can block pores and overlap particles. .

In the standard pore blocking model, when the particle size is much smaller than the pore size of the separator, the particles penetrate into the pores to adhere to the pore walls, blocking part or all of the pores to reduce pore volume . The reduced cedar bloom is proportional to the permeation volume.

Cake filtration model is that the particle size is larger than the pores of the separator and can not penetrate into the pores, and they are already blocking the pores or overlapping the particles existing on the surface to form a cake.

However, it is too complicated to identify only the actual membrane fouling mechanism because multiple membrane fouling mechanisms, including substantially one major mechanism, in membrane fouling occur simultaneously or sequentially. Therefore, in order to more accurately predict membrane fouling phenomena, it is necessary to consider the effect of combining multiple membrane fouling mechanisms and identify the main mechanisms at each stage of membrane fouling.

The present invention grasps the effect of a plurality of mechanisms integrally in actual membrane contamination and integrates four membrane contamination models into one expression to create a new model of membrane contamination integration, And provides a prediction method and a diagnostic method.

It also provides a method and system for predicting membrane fouling more precisely using a membrane fouling integration model and a major model that plays a leading role in membrane fouling among four membrane fouling models.

The membrane fouling predicting method using the integrated model according to the present invention is a method for predicting membrane fouling by using an integrated modeling unit to calculate the transmembrane pressure differential (TMP) using a full pore blocking model, a medium pore blocking model, a cake formation model and a standard pore blocking model ; The data collection unit periodically collecting actual data of membrane contamination including inter-membrane pressure differential data; Calculating the inter-membrane pressure difference of the integrated model; Calculating an optimization parameter for the integrated model using the inter-membrane pressure difference of the membrane-contaminated actual data and the inter-membrane pressure difference of the integrated model; And the film fouling predicting section calculates the inter-membrane pressure difference prediction value of the membrane contamination integrated model using the optimization parameter.

The step of generating the film contamination integrated model can be executed by the following equations (2) to (6).

&Quot; (2) "

(P / P ₀ : differential pressure across the membrane, R / R ₀ : permeation resistance, A / A ₀ :

&Quot; (3) "

(V: permeation volume of membrane integrated model t: time)

&Quot; (4) "

(R / R ₀ : permeation resistance of integrated membrane contamination model)

&Quot; (5) "

&Quot; (6) "

(P / P _0: transmembrane pressure difference of fouling integrated model, a: the first transmembrane pressure differential parameter, k _b: complete pore blocking parameters of the model variables, k _i: parameter of the median pore plugging model variables, k _c: cake formation model , K _s : parameter of standard pore clogging model, J ₀ : initial flux, t: time)

&Quot; (7) "

( _I : _measured : membrane-to-membrane pressure difference of membrane contamination measured data, _i : _model : membrane-to-membrane pressure difference of membrane contamination integrated model, m:

The step of calculating the inter-film pressure difference prediction value of the membrane contamination integrated model can be executed by substituting the optimization parameter calculated by the above equation (7) into the membrane contamination integrated model defined by the above equation (6).

The step of predicting the membrane fouling may further include determining the accuracy of the membrane contamination integrated model using the inter-membrane pressure differential value of the membrane fouling integrated model and the membrane-to-membrane differential pressure of the membrane-contaminated actual data.

The step of determining the accuracy of the membrane contamination integrated model may include calculating a root mean square deviation (RMSE) between a predicted value of the intermembral pressure differential pressure of the membrane contamination integrated model defined by the following equation (8) Error).

&Quot; (8) "

( _{i, modeled} : Prediction value of intermembrane pressure difference of membrane contamination integration model)

The membrane contamination diagnosis method using the integrated model is based on the integrated modeling section, the membrane contamination integrated model for calculating the TMP (Trans Membrane Pressure) using the full pore blockage model, the mesopore blockage model, the cake formation model and the standard pore blockage model ; The data collection unit periodically collecting actual data of membrane contamination including inter-membrane pressure differential data; Calculating the inter-membrane pressure difference of the integrated model; Periodically calculating an optimization parameter for the integrated model using the inter-membrane pressure difference of the membrane-contaminated actual data and the inter-membrane pressure difference of the integrated model; And diagnosing the model having the highest contribution in the membrane contamination phenomenon through a multiple regression analysis method in which the diagnostic unit determines the periodically calculated optimization parameter as an independent variable and the inter-membrane pressure difference of the membrane contamination measured data as a dependent variable do.

Wherein the step of diagnosing comprises performing a single regression analysis on the dependent variable using the full regression model to which the parameters of the complete pore clogging model, the intermediate pore clogging model, the cake formation model, and the standard pore clogging model are applied; Performing an F-test on the full regression model to calculate an average regression sum; Performing a single regression analysis using each of the reduced regression models except the parameters of the full pore clogging model, the intermediate pore clogging model, the cake formation model, and the standard pore clogging model; Performing an F-test on each of the reduced regression models to calculate an average regression sum; And diagnosing the model with the largest contribution using the difference between the average regression sum of the full regression model and the average regression sum of the reduced regression model.

A system for predicting membrane fouling using an integrated model according to the present invention includes: a data collection unit for periodically collecting actual data of membrane contamination including intermembrane pressure difference data; An integrated modeling unit for generating a membrane contamination integrated model for calculating the intermembrane pressure difference using a full pore blocking model, a middle pore blocking model, a cake formation model, and a standard pore blocking model; Calculating an inter-membrane pressure difference of the membrane contamination integrated model, calculating an optimization parameter for the membrane contamination integrated model using the inter-membrane pressure difference of the membrane contamination actual measurement data, And a film fouling predicting unit for calculating a predicted value of the pressure difference.

The film contamination predicting unit may determine the accuracy of the membrane contamination integrated model by using the inter-membrane pressure difference prediction value of the membrane contamination integrated model and the membrane-to-membrane pressure difference between the actual data of membrane contamination.

The membrane contamination diagnosis system using the integrated model according to the present invention includes: a data collection unit for periodically collecting actual data of membrane contamination including inter-membrane differential pressure data; An integrated modeling unit for generating a membrane contamination integrated model for calculating the intermembrane pressure difference using a full pore blocking model, a middle pore blocking model, a cake formation model, and a standard pore blocking model; A film contamination predicting unit for calculating an inter-membrane pressure difference of the membrane contamination integrated model and periodically calculating an optimization parameter for the membrane contamination integrated model using the inter-membrane pressure difference of the membrane contamination actual measurement data; And a diagnosing unit for diagnosing a model having the highest contribution in the membrane contamination phenomenon through a multiple regression analysis method in which the periodically calculated optimization parameter is used as an independent variable and the inter-membrane pressure difference of the membrane contamination measured data is used as a dependent variable .

The method and system for predicting and diagnosing membrane fouling using the integrated model of the present invention grasps the effect of a plurality of mechanisms integrally on actual membrane fouling and generates a new model of membrane fouling, Four types of membrane contamination models can be integrated and interpreted in a single equation.

In addition, membrane contamination phenomena can be more accurately predicted by using the membrane contamination integration model, and the main model that plays a leading role in membrane contamination among four membrane contamination models can be diagnosed.

1 is a schematic diagram of a static pressure plugging filtration model showing a decrease in permeation flux with time,
FIG. 2 is a block diagram of a film fouling prediction and diagnosis system using an integrated model according to an embodiment of the present invention. FIG.
3 is an operation flowchart of an integrated modeling unit according to an embodiment of the present invention,
4 is a flowchart illustrating an operation of a film contamination predicting unit according to an embodiment of the present invention.
5 is a graph showing the membrane contamination actual measured data, the membrane contamination integrated model and the intermembrane pressure difference of the individual membrane contamination models,
6 is a table that numerically calculates the mean square root deviation between measured data, the film contamination integrated model, and the inter-membrane pressure differential of the individual film contamination model,
7 is an operational flowchart of a diagnosis unit according to an embodiment of the present invention, and Fig.
8 is a table showing the contribution of the membrane fouling model according to an embodiment of the present invention.

Hereinafter, preferred embodiments of the present invention will be described with reference to the accompanying drawings. In the drawings, the same reference numerals are used to designate the same or similar components, and the same reference numerals will be used to designate the same or similar components. Detailed descriptions of known functions and configurations are omitted.

2 is a block diagram of a membrane fouling prediction and diagnosis system using an integrated model according to an embodiment of the present invention.

The data data collection unit 21 periodically collects actual data of the contamination of the membrane from the membrane filtration water treatment system 10. [ The membrane contamination actual measurement data includes data measured as a function of the membrane-to-membrane differential pressure of the separation membrane actually used in the membrane filtration water treatment system 10 according to a predetermined period. The data collecting unit 21 is connected to the membrane filtration and treatment system 10 in a wired manner or performs data communication using a wireless communication modem.

The integrated modeling unit 22 generates a mathematical model for integrally analyzing the effect of the complete pore blocking model, the intermediate pore blocking model, the cake forming model, and the standard pore blocking model on the separator. The mathematical steps for generating the membrane contamination integrated model of the integrated modeling unit 22 are as follows.

The permeation rate Q of the separation membrane can be expressed by Darcy's law expressed by the inter-membrane pressure P and the permeation resistance R as shown in the following equation (1). The permeation rate of the colloidal solution containing fine particles in the filtration process using the inter-membrane pressure difference P as a driving force is reduced over time. The phenomenon of the decrease in the permeation rate is due to the fact that the pores on the surface of the membrane are clogged with suspended particles or suspended particles in the solution accumulate on the surface of the membrane to form a cake layer.

[Equation 1]

In Equation (1), μ denotes an intrinsic viscosity, and A denotes an area of the separation membrane.

When the permeation rate is unchanged, the pressure difference tends to increase as the membrane fouling phenomenon of the separation membrane progresses, and [Equation 1] can be expressed by Equation (2).

&Quot; (2) "

(P / P ₀ : differential pressure across the membrane, R / R ₀ : permeation resistance, A / A ₀ :

In Equation (2), P / P ₀ means the change in the first intermembrane pressure difference, A ₀ / A means the change in the permeable area, and R / R ₀ means the increase in permeation resistance it means. In the following description, P / P ₀ is the inter-membrane pressure difference, R / R ₀ is the permeation resistance, and A / A ₀ is the permeation area.

When the membrane is contaminated by the complete pore blocking model and the intermediate pore blocking model, some of the pores of the membrane can not be filtered. Thus, during the filtration process, the flux in the unopened pores is increased.

The permeation area (A / A ₀ ) capable of filtration in the complete pore clogging model and the medium pore clogging model can be obtained from the following Table 1, and the numerical value calculated from [Table 1] 2], it is possible to calculate the increasing flux filtered through an area capable of filtering. In the following Table 1, k _b and k _i mean the parameters of the complete pore clogging model and the mesopore clogging model, J ₀ denotes the initial flux at the beginning of the filtration operation, t denotes the time Variable.

Complete pore clogging model Medium pore clogging model Transmission area (m ² )

Increasing flux (L / m ³ · hr)

Transmission area (m ² )

In the cake formation model and standard pore clogging model, the permeation resistance is increased by the cake layer formed on the surface of the separator, and the increased permeation resistance (R / R ₀ ) can be calculated using the following Table 2 .

Cake Formation Model Standard pore clogging model Permeation resistance

The membrane contamination integration model can be used to explain the decrease of the permeation area due to the clogging of the membrane and the increase of the permeation resistance due to the formation of the cake layer.

3 is an operation flowchart of the integrated modeling unit 22 according to an embodiment of the present invention.

When the clogging of the membrane occurs due to the combined effect of the complete pore clogging model and the medium pore clogging model, the integrated modeling unit 22 calculates the sum of the permeate volumes of the complete pore clogging model and the medium pore clogging model in Table 1 The permeation volume of the membrane contamination integration model can be calculated, which can be expressed by Equation (3) (S301).

&Quot; (3) "

(V: permeation volume of membrane contamination integrated model)

In addition, since the permeation resistance applied to the filtrate is generated by the combined effect of the cake forming model and the standard pore blocking model phenomenon, the integrated modeling unit 22 can calculate the permeation resistance The permeation resistance of the membrane contamination integrated model can be calculated (S302).

&Quot; (4) "

(R / R ₀ : permeation resistance of integrated membrane contamination model)

If the permeation volume V of the equation (3) is substituted into the equation (4), the permeation resistance of the membrane contamination integrated model can be expressed by the following equation (S303).

&Quot; (5) "

If the permeation resistance of the equation (5) is substituted into the equation (2), the equation for the membrane-contaminated integrated model transmembrane pressure as shown in the following equation (6) can be obtained (S304).

&Quot; (6) "

(P / P _0: transmembrane pressure difference of fouling integrated model, a: the first transmembrane pressure differential parameter, k _b: complete pore blocking parameters of the model variables, k _i: parameter of the median pore plugging model variables, k _c: cake formation model , K _s : parameter of standard pore clogging model, J ₀ : initial flux, t: time)

는 [표 1]의 투과면적(A/A₀)에 의하여 구할 수 있으며, a는 막 오염의 최초 값과 일치시키기 위한 매개변수이다.In Equation (6)

Can be obtained by the permeation area (A / A ₀ ) of [Table 1], and a is a parameter for matching the initial value of the membrane contamination.

4 is a flowchart of the operation of the film contamination predicting unit. The film contamination predicting unit 23 calculates four parameters k _b , k _i , k _c, and k _s of the membrane contamination integration model represented by Equation (6) using the actual data of film contamination stored in the database, (GA, Genetic Algorithm). Genetic algorithms can perform optimizations in terms of the parallel search of parameters without differential computation of the objective function or special mathematical operations.

The film contamination predicting unit 23 can calculate a set of optimization parameters for four parameters by applying a least squares error method to the inter-membrane pressure difference of the film contamination integrated model and the membrane contamination actual measured data, Can be expressed by an objective function expressed by Equation (7) (S401).

&Quot; (7) "

( _I : _measured : membrane-to-membrane pressure difference of membrane contamination measured data, _i : _model : membrane-to-membrane pressure difference of membrane contamination integrated model, m:

In Equation (7), _{I, measure} represents the inter-membrane pressure difference of the actual data of membrane contamination, _{i, model} represents the inter-membrane pressure difference of the membrane contamination integrated model, and m represents the number of data of the membrane contamination actual data.

The film contamination predicting unit 23 substitutes the optimization parameters for the four parameters calculated using Equation (7) into Equation (6) to calculate the film contamination integration model (Step S402).

The film contamination predicting unit 23 calculates the film contamination predicted value by using the mean square root error (RMSE) between the inter-film pressure difference prediction value of the film contamination integrated model and the inter-membrane pressure difference of the film contamination actual data as shown in the following equation (8) The accuracy of the inter-membrane pressure difference predicted value of the contaminated integrated model can be evaluated (S403). The film contamination predicting unit 23 can judge that the film contamination integrated model is closer to the actual model as the value of the mean square root deviation calculated by the following formula (8) becomes smaller.

&Quot; (8) "

( _{i, modeled} : Prediction value of intermembrane pressure difference of membrane contamination integration model)

In Equation (8), _{i, modeled} represents a predictive value of the inter-membrane pressure difference of the membrane contamination integrated model.

FIGS. 5A and 5B are graphs showing predicted intermembrane pressure differentials and predicted intermembrane pressure differentials using the predicted intermembrane pressure differentials and membrane fouling models of two membrane contaminated actual data, and FIG. This is a table that numerically calculates the average square root deviation of the inter-membrane pressure differential versus membrane contaminant integrated model of the contaminated actual data and the predicted value of inter-membrane pressure difference of the individual membrane contamination model. It can be seen that the mean square deviation of the predicted value of the intermembrane pressure differential of the membrane contamination integrated model is smaller than that of the lowest predicted value of the intermembrane pressure difference of the individual membrane contamination model.

The diagnosis unit 24 can evaluate the contribution of each model through a multiple regression analysis on the measured data on membrane contamination which periodically collects the optimization parameters calculated using the membrane contamination integration model in the data collection unit 21. [

For example, if the data collection unit 21 collects actual data of film contamination at intervals of two days for 50 days, the film contamination predicting unit 23 calculates the optimization parameters for the first 2 days (i.e., m = 2) And finally calculates a set of optimization parameters according to the number of data for 50 days (i.e., m = 25). Therefore, the optimization parameter set is periodically updated in accordance with the actual data collection period of the data acquisition unit 21.

The diagnostic unit 24 applies a multiple regression analysis (MRA) to the set of optimization parameters that are periodically updated to find the model with the highest contribution among the four models in the membrane contamination phenomenon. The diagnosis unit 24 sets 25 optimization parameter sets k _b , k _i , k _c, and k _s updated at intervals of 2 days for 50 days as an independent variable, for example _{, Set the measure} as a dependent variable.

The diagnosis unit 24 performs a regression analysis on the dependent variable using a full regression model in which all four independent variables are applied (S701), performs an F-test on the dependent variable, The mean squared regression (MSR) for the model is calculated (S702). The mean regression fitting for the full regression model is an index of how well the full regression model explains the intermembrane pressure of the membrane contamination data.

The diagnostic unit 24 then performs a regression analysis on the dependent variable using each reduced regression model except for one of the four independent variables (S703), and performs an F-test on the dependent variable The average regression sum for the reduced regression model is calculated (S704). (K _b , k _i , k _c ), (k _b , k _{i ,} k _s ) and (k _b , k _{c ,} k _s ) _, respectively, since the independent variable of one of the four independent variables is excluded. And (k _i , k _{c ,} k _s ) may be used as the independent variable, so that the diagnosis unit 24 calculates the average regression sum for the four reduced regression models.

The diagnostic unit 24 uses the difference between the mean regression sum of the four regression regression models and the mean sum sum for the full regression model so that each independent variable, that is, each optimization parameter contributes to the film contamination measured data (S705). The diagnostic unit 24 determines whether the difference between the mean regression sum for the four regression regression models and the mean difference sum for the full regression model is the largest, Can be evaluated as a parameter (S706).

Figure 8 shows the contribution and contribution of each of the four parameters. In the experimental data 1, the difference between the average regression sum of the regression regression model excluding the parameter (k _i ) of the mesopore clogging model and the average regression sum of the full regression model was the largest. In the experimental data 2, It can be seen that the difference between the average regression sum of the regression model with the variable (k _c ) excluded and the average regression sum of the regression model is the largest. As a result, the mesopore clogging model in the actual data 1 and the cake formation model in the actual data 2 can be regarded as parameters having a large contribution to the membrane contamination.

Embodiments of the present invention include computer readable media including program instructions for performing various computer implemented operations. The computer-readable medium may include program instructions, data files, data structures, and the like, alone or in combination. The program instructions recorded on the medium may be those specially designed and constructed for the present invention or may be available to those skilled in the art of computer software. Examples of computer-readable media include magnetic media such as hard disks, floppy disks and magnetic tape, optical media such as CDROMs, DVDs, magneto-optical media such as floptical disks, Magneto-optical media, and hardware devices specifically configured to store and perform program instructions such as ROM, RAM, flash memory, and the like. Examples of program instructions include machine language code such as those produced by a compiler, as well as high-level language code that can be executed by a computer using an interpreter or the like.

As described above, the present invention has been described with reference to particular embodiments, such as specific elements, and specific embodiments and drawings. However, it should be understood that the present invention is not limited to the above- And various modifications and changes may be made thereto by those skilled in the art to which the present invention pertains. Therefore, the spirit of the present invention should not be construed as being limited to the described embodiments, and all of the equivalents or equivalents of the claims, as well as the following claims, are included in the scope of the present invention.

10: membrane filtration water treatment system
20: Pollution Prediction and Diagnosis System
21: Data collecting unit
22: Integrated Modeling Unit
23: membrane pollution predicting unit
24:

Claims (11)

Creating an integrated membrane fouling model for calculating the intermembrane pressure differential (TMP) using an integrated modeling section full pore clogging model, a mesopore clogging model, a cake formation model, and a standard pore clogging model;
The data collection unit periodically collecting actual data of membrane contamination including inter-membrane pressure differential data;
Calculating a membrane-to-membrane differential pressure of the integrated model;
Periodically calculating an optimization parameter for the integrated model using the inter-membrane pressure difference of the membrane-contaminated actual data and the inter-membrane pressure difference of the integrated model;
And diagnosing the model having the highest contribution in the membrane contamination phenomenon through a multiple regression analysis method in which the diagnostic unit determines the periodically calculated optimization parameter as an independent variable and the inter-membrane pressure difference of the membrane contamination measured data as a dependent variable In addition,
Wherein the step of diagnosing comprises performing a single regression analysis on the dependent variable using the full regression model to which the parameters of the complete pore clogging model, the intermediate pore clogging model, the cake formation model, and the standard pore clogging model are applied; Performing an F-test on the full regression model to calculate an average regression sum; Performing a single regression analysis using each of the reduced regression models except the parameters of the full pore clogging model, the intermediate pore clogging model, the cake formation model, and the standard pore clogging model; Performing an F-test on each of the reduced regression models to calculate an average regression sum; And diagnosing a model with the largest contribution using a difference between an average regression sum of the full regression model and an average regression sum of the reduced regression model.
The method according to claim 1,
Wherein the step of generating the film contamination integrated model is performed by the following equations (2) to (6).
&Quot; (2) "

(P / P ₀ : differential pressure across the membrane, R / R ₀ : permeation resistance, A / A ₀ :
&Quot; (3) "

(V: permeation volume of membrane integrated model t: time)
&Quot; (4) "

(R / R ₀ : permeation resistance of integrated membrane contamination model)
&Quot; (5) "

&Quot; (6) "

(P / P _0: transmembrane pressure difference of fouling integrated model, a: the first transmembrane pressure differential parameter, k _b: complete pore blocking parameters of the model variables, k _i: parameter of the median pore plugging model variables, k _c: cake formation model , K _s : parameter of standard pore clogging model, J ₀ : initial flux, t: time)
3. The method of claim 2, wherein calculating the optimization parameter comprises:
A method for diagnosing membrane fouling using an integrated model executed by an objective function (f) defined by the following equation (7) using a genetic algorithm.
&Quot; (7) "

( _{I, measured} : membrane-to-membrane pressure difference of membrane contamination measured data, _{i, model} : membrane-to-membrane pressure difference of membrane contamination combined model, m:
4. The method of claim 3, wherein the step of calculating the inter-
And substituting the optimization parameter calculated by Equation (7) into the membrane contamination integration model defined by Equation (6) above.
5. The method of claim 4,
Determining the accuracy of the membrane contamination integrated model by using the predicted membrane fouling pressure differential value of the membrane fouling integrated model and the membrane-to-membrane differential pressure of the membrane contaminated actual data
A method for diagnosing membrane fouling using an integrated model.
6. The method of claim 5, wherein determining the accuracy of the membrane contamination integration model comprises:
(RMSE) between the predicted value of the intermembral pressure differential pressure of the integrated membrane contamination model defined by the following equation (8) and the intermembrane pressure difference of the actual data of membrane contamination using the integrated model Method of diagnosing membrane fouling.
&Quot; (8) "

( _{i, modeled} : Prediction value of intermembrane pressure difference of membrane contamination integration model)
delete delete A data collecting unit for periodically collecting actual data of membrane contamination including inter-membrane differential pressure data;
An integrated modeling unit for generating a membrane contamination integrated model for calculating the intermembrane pressure difference using a full pore blocking model, a middle pore blocking model, a cake formation model, and a standard pore blocking model;
A film contamination predicting unit for calculating an inter-membrane pressure difference of the membrane contamination integrated model and periodically calculating an optimization parameter for the membrane contamination integrated model using the inter-membrane pressure difference of the membrane contamination actual measurement data; And
And a diagnosis unit for diagnosing a model having the highest contribution in the membrane contamination phenomenon through a multiple regression analysis method in which the periodically calculated optimization parameter is used as an independent variable and the inter-membrane pressure difference of the membrane contamination measured data is used as a dependent variable,
The diagnostic unit performs a single regression analysis on the dependent variable using the full regression model to which the parameters of the complete pore clogging model, the intermediate pore clogging model, the cake formation model, and the standard pore blockage model are applied, To calculate an average regression coefficient, and using each of the reduced regression models except for the parameters of the full pore clogging model, the intermediate pore clogging model, the cake formation model, and the standard pore clogging model, Performing a single regression analysis and performing an F-test on each of the reduced regression models to calculate a mean regression sum, and calculating a difference between an average regression sum of the full regression model and an average regression sum of the reduced regression model A Membrane Fouling Diagnosis System Using an Integrated Model to Diagnose the Model with the Largest Contribution.
10. The method of claim 9,
Wherein the film fouling predicting unit estimates the accuracy of the membrane contamination integrated model by using the inter-membrane pressure difference estimated value of the membrane fouling integrated model and the membrane-to-membrane differential pressure of the membrane contaminated actual data. delete

Images