KR20160114211A - Wastewater treatment systeme and method - Google Patents

Abstract

The present invention relates to a water treatment system and a method. The present invention provides an economically efficient water treatment system and a method which not only stably treat water, but also reduce energy. The water treatment system comprises: a water quality data collection unit; a power data collection unit; and a water treatment management system.

Description

TECHNICAL FIELD [0001] The present invention relates to a water treatment system,

The present invention relates to a water treatment system and method, and more particularly, to a water treatment system and method in which an energy optimization operation technique is introduced efficiently.

Wastewater treatment plant is a biological process and its reaction is very complicated, which makes it difficult to understand the process behavior and also requires a lot of analysis to know the process condition and behavior. Accordingly, the operation of almost all sewage treatment plants now depends on the experience knowledge of skilled drivers. This is not desirable from the viewpoint of automation of water treatment process and reduction of labor cost. In addition, driver's empirical knowledge is reliable when the accumulated amount of knowledge is large, but it is difficult to guarantee the permanence, objectivity and amount of information that can be processed. In addition, when operating the sewage treatment plant depending on the driver's empirical knowledge, there is a great difficulty in reducing the energy of the sewage treatment plant.

The background art of the present invention is disclosed in Korean Patent Laid-Open Publication No. 2006-0095273 (published on Aug. 31, 2006, integrated energy circulation utilization system linking a sewage treatment system and a cogeneration system).

The present invention provides an economically efficient water treatment system and method that reduces energy as well as stable water treatment.

The objects of the present invention are not limited to the above-mentioned objects, and other objects not mentioned can be clearly understood from the following description.

According to an aspect of the present invention, a water treatment system is provided.

The water treatment system according to an embodiment of the present invention includes a water quality data collection unit for collecting water quality data in each water treatment process, a power data collection unit for collecting power data from the power equipment installed in each water treatment process, And a water treatment management system for setting an optimal water treatment control method according to operating costs calculated based on operating factors and power consumption that satisfy the water quality standard range using the data.

According to another aspect of the present invention, a water treatment method and a computer program for executing the water treatment method are provided.

The water treatment method and the computer program for executing the water treatment method according to an embodiment of the present invention include collecting operational data including water quality data of each water treatment process and power data of electric power equipment used in each water treatment process, Deriving operating parameters of the water treatment process utilized to achieve water quality standards and water quality standards, calculating the range of operating factors satisfying the range of water quality standards using water quality data and power data, And the water quality data of the optimum model is selected as a variable to select an optimum model for analyzing the correlation between the operation factor and the treated water quality, a step of analyzing the power consumption according to the operation parameter of the optimum model, And a step of setting an optimal control method of the water treatment process based on the operation cost It may contain.

The present invention enables smart operation diagnosis based on the field through efficient management of water treatment system.

The present invention enables the establishment of an energy optimization operation technology of a water treatment system by analyzing the correlation between the main operating factors, the treated water quality, and the energy efficiency and suggesting the optimum operating factors.

The present invention proposes maintenance cost and energy saving plan through smart operation of sewage and wastewater treatment facility and it is possible to develop standardization technology of process diagnosis.

1 is a view for explaining a water treatment system according to an embodiment of the present invention;
2 is a configuration diagram of a water treatment system according to an embodiment of the present invention;
3 is a view for explaining a water treatment method according to an embodiment of the present invention;
4 to 15 are views for explaining the water treatment method according to an embodiment of the present invention in more detail.

Hereinafter, embodiments of the present invention will be described in detail with reference to the accompanying drawings so that those skilled in the art can easily carry out the present invention. The present invention may, however, be embodied in many different forms and should not be construed as limited to the embodiments set forth herein. Also, when a part is referred to as "including " an element, it does not exclude other elements unless specifically stated otherwise.

DETAILED DESCRIPTION OF THE PREFERRED EMBODIMENTS Reference will now be made in detail to embodiments of the present invention, examples of which are illustrated in the accompanying drawings.

1 is a view for explaining a water treatment system according to an embodiment of the present invention.

Referring to FIG. 1, a water treatment system according to the present invention includes a water treatment management system 100, a water quality data collection device 200, and a power data collection device 300.

The water treatment management system 100 inputs the water quality data in each water treatment process and inputs the power data from the electric power equipment installed in the water treatment process. The water treatment management system 100 calculates the range of the water treatment process operation factors satisfying the water quality standard range using the collected water quality data and the power data and calculates an optimum model for the correlation analysis between the operation factor and the treated water quality . The water treatment management system 100 analyzes power usage according to a range of operational factors of the selected optimal model, and calculates an economically efficient operation cost based on operating factors and power consumption. The water treatment management system 100 sets an optimum water treatment control method based on economical operating costs and controls each water treatment process. This will be described in more detail in Fig.

The water quality data collection device 200 is installed in each water treatment process to collect water quality data. Here, the water treatment process may include at least one of a primary settling tank, an anaerobic tank, an anoxic tank, an aerobic tank, and a secondary settling tank, but is not limited thereto. In addition, the water quality data is data representing the water quality, for example, temperature, pH, dissolved oxygen amount (DO), biological oxygen demand amount (BOD), chemical oxygen demand amount (COD), sludge amount (SS) Nitrogen (TN), and the like. The water quality data collection device 200 is connected to the water treatment management system 100 through wired / wireless lines through a network and transmits collected water quality data to the water treatment management system 100. Here, the water quality data can be converted in data format for transmission to the water treatment management system 100.

The power data collection device 300 collects power data representing the power consumption of the process facility from the power consumption measurement device connected to the process facility of each water treatment process. Here, the process equipment includes, but is not limited to, for example, an inflow pump, a blower, a motor, and the like. Here, the power data can be converted to a data format for transmission to the water treatment management system 100.

2 is a configuration diagram of a water treatment system according to an embodiment of the present invention.

2, the water treatment management system 100 includes an operation data collection unit 110, an operation factor derivation unit 120, an optimal model selection unit 130, a power consumption analysis unit 140, (150) and an optimal control method setting unit (160).

The operation data collection unit 110 collects operational data in each water treatment process. Here, the operational data includes the water quality data collected in each water treatment process and the power data collected from the power equipment installed in each water treatment process. The water treatment process includes all of the processes of applying at least one of physical, chemical, and biological treatments for improving water quality such as a sewage treatment process, a water treatment process, and a wastewater treatment process. Power devices include process equipment that uses electrical power, such as, for example, input pumps, blowers, and the like.

The operating factor deriving unit 120 derives operational factors for operating the water treatment process and calculates the range of the water treatment process operating factors satisfying the water quality standard range using the collected operational data. Herein, the operational factors include factors related to the operation of the water treatment management system 100 such as F / M (Food to Microorganism) ratio, Solids Retention Time (SRT), Carbon to Nitrogen And may include at least one of Carbon to Phosphorus (P) ratio, Hydraulic Retention Time (HRT), Specific Nitrification Rate (SNR), Specific Denitrification Rate (SDNR), and Dissolved Oxygen (DO).

The optimal model selection unit 130 selects an optimal model for analyzing the correlation between the operation factor and the treated water quality. Here, the optimal model simulates at least two of a linear regression model, a multiple regression model, a nonlinear probabilistic model, an artificial neural network model, and a deterministic model, ASM (Activated Sludge Models) Can be selected. The optimal model selection unit 130 analyzes a multiple regression analysis model using a linear probabilistic model and, if the prediction accuracy of a plurality of multiple regression analysis models generated by setting different conditions is within a preset allowable range, One regression model with multiple regression coefficients can be selected for each regression model. The optimal model selection unit 130 analyzes an artificial neural network model using a nonlinear probabilistic model and determines the predictive accuracy of a plurality of artificial neural network models generated by setting different parameters for data learning of the artificial neural network model within a predetermined allowable range , One artificial neural network model with a high sensitivity of the controllable operational factors calculated for each of the plurality of artificial neural network models can be selected. The optimal model selection unit 130 analyzes the ASM (Activated Sludge Models) model as a deterministic model, and when the prediction accuracy of the plurality of ASM models generated by setting the reaction formula variable of the ASM model to within a predetermined allowable range, One ASM model with a high sensitivity of the controllable operational factors calculated for each of the ASM models can be selected. Thereafter, the optimal model selection unit 130 can finally select one of the multiple regression analysis models, artificial neural network models, and ASM models having the highest prediction accuracy, respectively.

The power usage analysis unit 140 analyzes power usage according to a range of operation factors of the selected optimal model.

The economic efficiency analysis unit 150 calculates an economically efficient operation cost based on the analyzed operating factors and power consumption. Here, the economically efficient operation cost may include at least one of a drug cost, a power use cost, a facility maintenance cost, and a labor cost based on a range of operating factors of a water treatment process. Here, the electric power usage cost can be changed according to at least one of time period, season, and power peak.

The optimal control method setting unit 160 sets an optimal control method of the water treatment process according to the economic efficiency analysis.

3 is a view for explaining a water treatment method according to an embodiment of the present invention.

Referring to FIG. 3, in step S310, the water treatment management system 100 collects operational data including water quality data of each water treatment process and power data of power devices used in each water treatment process.

In step S320, the water treatment management system 100 derives the operating factors of the water treatment process utilized to achieve the predetermined water quality standards and water quality standards.

In step S330, the water treatment management system 100 calculates the range of water treatment process operation factors satisfying the water quality reference range using the collected water quality data and power data.

In step S340, the water treatment management system 100 sets the obtained operating parameters and the water quality data of the effluent water as variables, respectively, and selects an optimal model for analyzing the correlation between the operation factors and the treated water quality. This will be described in detail later with reference to FIG. 4 to FIG.

In step S350, the water treatment management system 100 analyzes the power consumption according to the operating parameter range of the optimum model water treatment process.

In step S360, the water treatment management system 100 calculates an economically efficient operation cost based on the water treatment process operation factor and power consumption. For example, the water treatment management system 100 can accurately calculate the electric power charges for each operation factor by searching the electric power charges calculated for each day and hour according to the electric power usage amount for each operation factor.

In step S370, the water treatment management system 100 sets an optimal control method of the water treatment process based on the economical operating cost.

FIGS. 4 to 15 are views for explaining the optimal model selection method according to an embodiment of the present invention in more detail.

Referring to FIG. 4, in step S342, the water treatment management system 100 selects the derived operating factors as independent variables, and sets the set water quality standards as dependent variables. Here, the water quality standard can use the water quality data of the effluent.

In step S344, the water treatment management system 100 generates a plurality of models for simulating the correlation analysis of the operation factors and the treated water quality. Here, the plurality of models may include at least two of a linear regression model, a multiple regression analysis model, a nonlinear stochastic model, an artificial neural network model, and a deterministic model, ASM (Activated Sludge Models) model.

Hereinafter, the generation of the multiple regression analysis model will be described in detail in order.

Referring to FIG. 5, in step S510, the water treatment management system 100 inputs a selected independent variable (operation factor) and a dependent variable (a set water quality standard) to the multiple regression analysis model.

8, the water treatment management system 100 sets three variables such as the effluent COD 810, the effluent TN 820, and the effluent TP 830 as dependent variables, for example, as the treated water quality data, You can set 28 variables as independent variables.

In step S520, the water treatment management system 100 executes a step selection method multiple regression analysis. Here, the multiple regression analysis is a statistical technique for analyzing the relationship between the dependent variable (Y) and the independent variables (X1, X2, X3 ...... Xn). The general form of the multiple linear regression model is expressed as:

Equation 1

: 종속변수

: Dependent variable

: 부분회귀계수

: Partial regression coefficient

: 독립변수

: Independent variable

: 오차항

: Error Harbor

There are three types of variable selection methods for multiple regression models: Forward selection, Backward elimination, and Stepwise selection. The forward selection method is a method to find the most appropriate model by increasing one by one from the independent variable having the greatest influence on the dependent variable. The backward elimination method includes all the independent variables that are set, and then removes the insignificant independent variables one by one. How to find it. These two variable selection methods are disadvantageous in that the selected independent variable is not removed from the model. One of the ways to overcome this drawback is step selection. The step selection method is similar to the forward selection method, but each time the independent variable is included in the model, the independent variables that have already been selected are subjected to a significant test to determine whether the independent variable remains. The water treatment management system 100 can use statistical analysis software SAS 9.2 for multiple regression analysis and can set the significance level to 0.05 when selecting the F test of the model and the variable selection method.

In step S530, the water treatment management system 100 removes the independent variable after the examination if the multi-collinearity problem between the independent variables occurs in the regression analysis process. In this case, the multi-collinearity is evaluated by selecting the variable using the step selection method and then calculating the variance inflation factor (VIF) of the variable. If there is a variable having a value of 10 or more, Again, use the step selection method to select a variable.

Thereafter, the water treatment management system 100 identifies the standardized regression coefficient by classifying into CASE 1 and CASE 2 according to whether or not the abnormal value and the influence value are applied. Here, CASE 1 does not use the test for outliers and influence values, while CASE 2 uses tests for outliers and influences.

In step S540, the water treatment management system 100 removes the outliers and the influence values and calculates the standardized regression coefficient in order to disable the test for the abnormal value and the influence value (CASE 1).

In step S550, the water treatment management system 100 calculates a standardized regression coefficient by applying a test on the outliers and the influence values (CASE 2).

Referring to FIG. 9, the water treatment management system 100 may calculate a standardized regression coefficient for an operational factor, for example, as shown in the table of FIG. The water treatment management system 100 can compare the standardized regression coefficients for the three dependent variable effluent COD 910, the effluent TN 920, and the effluent TP 930 to list independent variables in order of greater influence, It is possible to classify operating factors that can be controlled midway separately.

In steps S560 and S570, the water treatment management system 100 calculates RMSE and MAPE of CASE 1 and CASE 2, respectively. Here, RMSE (Root Mean Square Error) is an intuitive and meaningful estimation method of accuracy of the prediction model, and RMSE is calculated by the following equation (2).

Equation 2

은 관측치의 수,

는 실측치 및

는 모델 예측치임.here,

Is the number of observations,

And

Is the model predicted value.

The RMSE is the mean square root of the residual, which is the difference value, and the smaller the RMSE, the more accurate the model is.

Here, MAPE (Mean Absolute Percentage Error) is calculated by the following equation (3).

Equation 3

는 예측 자료의

번째 관측값,

는 예측 자료의

번째 예측값,

은 예측 자료의 수임.here

Of the forecast data

The second observation,

Of the forecast data

Th prediction value,

Is the number of forecasts.

MAPE can be interpreted as highly accurate prediction if MAPE <10, excellent prediction if 10 ≤ MAPE <20, rational prediction if 20 ≤ MAPE <50, and inaccurate prediction if MAPE> 50.

10, the water treatment management system 100 may calculate RMSE and MAPE for three dependent variables such as effluent COD 1010, effluent T-N 1020, and effluent T-P 1030.

In step S565 and step S575, the water treatment management system 100 determines whether the RMSE and the MAPE of each of CASE 1 and CASE 2 are within the allowable range set in advance.

In step S580, the water treatment management system 100 finally selects a model having a higher sum of standardized regression coefficients of the controllable operating factors of the model when the RMSE and the MAPE are within the predetermined allowable ranges, respectively. This is because the higher the sum of the standardized regression coefficients, the higher the possibility of reducing the power consumption by the driver's choice. For example, when the allowable range of the RMSE error is set to 10% or less, the water treatment process management system 100 may set the error of CASE 1 to 8.5% and the error of CASE 2 to 9.0% A model with a higher sum of regression coefficients can be chosen. For example, when the error of the CASE 1 is 8.5% and the error of the CASE 2 is 13%, the water treatment management system 100 can select the CASE 1 model because the error of the CASE 2 deviates from the preset allowable range have.

Hereinafter, generation of an artificial neural network model will be described in detail.

Referring to FIG. 6, in step S610, the water treatment management system 100 inputs a selected independent variable (operation factor) and a dependent variable (a set water quality standard) to the artificial neural network model. Here, the artificial neural network model is a mathematical model that simplifies the neurotransmission process of a living organism. It is a model in which neurons (neurons - neurons: the smallest units forming a network in a neural network) The problem is analyzed through a kind of training process that adjusts. This process is similar to the process of learning and memorizing by humans, through which reasoning, classification, and prediction can be performed.

In step S620, the water treatment management system 100 sets parameters for learning of the artificial neural network data. Here, the parameter may include at least one of the number of hidden layers, a threshold value, an algorithm to be used, and an error value, but is not limited thereto.

In step S630 and step S635, the water treatment management system 100 performs the first artificial neural network model data learning and the second artificial neural network model data learning in which parameters are set differently, respectively. Here, the first artificial neural network model and the second artificial neural network model represent a plurality of artificial neural network models, and may be two or more artificial neural networks having different parameters. 11 is an exemplary diagram of an artificial neural network learning page according to an embodiment of the present invention. Artificial neural network learning data is organized by learning execution date, and consists of a creation date, a data description, an input variable, and a list of output variables. The selected neural network learning data consists of input variables and output variables. After inputting the input values to be simulated in the input variables, simulation values of the output variables are returned from the neural network server when the operation button is clicked.

In steps S640 and S645, the water treatment management system 100 calculates RMSE and MAPE for the first and second artificial neural network models, respectively.

12, the water treatment management system 100 calculates three parameters, for example, the effluent COD 1110, the effluent TN 1120, and the effluent TP 1130 as the treated water quality data, MAPE can be calculated.

In step S650 and step S655, the water treatment management system 100 analyzes the sensitivity of each of the first and second artificial neural network models by operating factors. Here, the controllable operating factors can be classified separately.

In step S660 and step S665, the water treatment management system 100 determines whether the RMSE and the MAPE of each of the first artificial neural network model and the second artificial neural network model are within a predetermined allowable range. For example, when the allowable range of the RMSE error is set to 10% or less, the water treatment management system 100 sets the first error rate of the first artificial neural network model to 14.5% and the error of the second artificial neural network model to 9.5% It can be excluded because the error of the artificial neural network model deviates from the preset allowable range.

In step S670, the water treatment management system 100 finally selects an artificial neural network model having a high sensitivity of the controllable operational factors of the model when the RMSE and the MAPE are within the predetermined allowable range, respectively. For example, if the error of the first artificial neural network model is 7.5% and the error of the second artificial neural network model is 9.5%, then the water treatment management system 100 can detect the sensitivity of the controllable operating factors You can choose a higher model.

Hereinafter, generation of the ASM model will be described in detail.

Referring to FIG. 7, in step S710, the water treatment management system 100 inputs a dependent variable (based on the set water quality) to the ASM model. Here, the ASM model may be a simplified ASM model in which some parameters are omitted.

In step S720, the water treatment management system 100 determines the reaction formula of the target substance in each reaction tank.

In step S730 and step S735, the water treatment management system 100 optimizes the reaction parameters of the first ASM model and the second ASM model in which the reaction equation variable is set differently, and optimizes the model. Referring to FIGS. 13 to 15, the first ASM model and the second ASM model can be optimized by setting reaction parameters and parameters, respectively.

In steps S740 and S745, the water treatment management system 100 calculates the RMSE and the MAPE of each of the first ASM model and the second ASM model.

In step S750 and step S755, the water treatment management system 100 analyzes the sensitivity of each of the first ASM model and the second ASM model according to the operating factors through the set operating factor assignment condition modeling. Here, the controllable operating factors can be classified separately.

In steps S760 and S765, the water treatment management system 100 determines whether the RMSE and the MAPE of each of the first ASM model and the second ASM model are within a predetermined allowable range. For example, when the allowable range of the RMSE error is set to 10% or less and the error of the first ASM model is 8.5% and the error of the second ASM model is 13.0% The error can be excluded because the error is out of the preset allowable range.

In step S770, the water treatment management system 100 finally selects the ASM model having high sensitivity of the controllable operation factors of the model when the RMSE and the MAPE are within the predetermined allowable ranges, respectively. For example, if the error of the first ASM is 8.5% and the error of the second ASM model is 9.0%, the water treatment management system 100 can obtain a model with high sensitivity of the controllable operation factors You can choose.

In step S344, the water treatment management system 100 selects an optimal model based on the prediction accuracy of each model. Here, the prediction accuracy may be based on at least one of RMSE and MAPE.

The water treatment methods according to various embodiments of the present invention may be implemented in the form of program instructions that can be executed through various means such as servers. Further, a program for executing the water treatment method according to the present invention may be installed in a computer means and recorded in a computer-readable medium. The computer readable medium may include program instructions, data files, data structures, and the like, alone or in combination.

Program instructions to be recorded on a computer-readable medium may be those specially designed and constructed for the present invention or may be available to those skilled in the computer software arts. Examples of computer-readable media include magnetic media such as hard disks, floppy disks, and magnetic tape; optical media such as CD-ROMs and DVDs; magnetic media such as floppy disks; Includes hardware devices specifically configured to store and execute program instructions such as magneto-optical media and ROM, RAM, flash memory, and the like.

The embodiments of the present invention have been described above. It will be understood by those skilled in the art that the foregoing description of the present invention has been presented for illustrative purposes and that those skilled in the art will readily understand that various changes in form and details may be made therein without departing from the spirit and scope of the invention. It will be possible. It is therefore to be understood that the above-described embodiments are illustrative in all aspects and not restrictive. The scope of the present invention is defined by the appended claims rather than the detailed description, and all changes or modifications derived from the meaning and scope of the claims and their equivalents should be construed as being included in the scope of the present invention.

100: water treatment management system
200: Water quality data collection device
300: Power data acquisition device

Claims (19)

In a water treatment system,
A water quality data collection unit for collecting water quality data in each water treatment process;
A power data collection unit for collecting power data from the power devices installed in the respective water treatment processes;
And a water treatment management system for setting an optimal water treatment control method according to operating costs calculated based on operation factors and power consumption amounts satisfying a water quality reference range using the water quality data and the power data.
The method according to claim 1,
The water treatment management system
An operation data collection unit for collecting operation data including the water quality data and the power data;
An operational factor deriving unit for calculating a range of operational factors satisfying the water quality standard range using the collected operational data;
An optimal model for selecting an optimal model for analyzing a correlation between the operation factor and the treated water quality;
A power usage analyzer for analyzing power usage according to a range of operating factors of the optimal model;
An economic efficiency analyzer for calculating operating costs based on the operating factors and the power usage; And
And an optimal control method setting unit for setting an optimum control method of the water treatment process according to the operation cost.
3. The method of claim 2,
The optimal model selection unit
And simulating at least two of a linear probabilistic model, a non-linear probabilistic model, and a deterministic model to select the optimal model.
The method of claim 3,
Wherein the linear probability model, the nonlinear probabilistic model, and the deterministic model are selected from a plurality of models having a high prediction accuracy.
5. The method of claim 4,
The linear probabilistic model is a multiple regression analysis model,
When the prediction accuracy of a plurality of multiple regression analysis models generated by setting different conditions is within a preset allowable range, a sum of standardized regression coefficients of controllable operational factors calculated for each of the plurality of multiple regression analysis models is high A water treatment system that selects multiple regression analysis models.
5. The method of claim 4,
The non-linear probabilistic model is an artificial neural network model,
When the prediction accuracy of a plurality of artificial neural network models generated by setting parameters for data learning of the artificial neural network model is within a predetermined allowable range, the sensitivity of the controllable operational factors calculated for each of the plurality of artificial neural network models A water treatment system for selecting a single artificial neural network model.
5. The method of claim 4,
The deterministic model is an ASM (Activated Sludge Models) model,
When a prediction accuracy of a plurality of ASM models generated by setting reaction formulas of the ASM model differently is within a predetermined allowable range, a single ASM model having a high sensitivity of the controllable operational factors calculated for each of the plurality of ASM models Water treatment system.
3. The method of claim 2,
Wherein the operating cost comprises at least one of a drug cost, a power usage cost, a facility maintenance cost, and a labor cost based on a range of operational factors of the respective water treatment processes.
9. The method of claim 8,
Wherein the power usage cost is calculated by using a power rate calculated by day and time according to power usage amount for each operation factor.
In the water treatment method,
Collecting operational data including water quality data of each water treatment process and power data of the power equipment used in each water treatment process;
Deriving operational parameters of a water treatment process utilized to achieve a predetermined water quality standard and the water quality standard;
Calculating a range of the operational factors satisfying the range of the water quality standard using the water quality data and the power data;
Selecting an optimal model for analyzing the correlation between the operational factors and the treated water quality by setting the operational parameters and the water quality data of the effluent water as variables, respectively;
Analyzing power usage according to an operation parameter of the optimal model;
Calculating operating costs based on the operating factors and the power usage; And
And setting an optimal control method of the water treatment process based on the operating cost.
11. The method of claim 10,
And selecting the optimum model for analyzing the correlation between the operation factor and the treated water quality by setting the operational parameter and the water quality data of the effluent water as variables, respectively
Selecting the derived operating factor as an independent variable and selecting the water quality data of the effluent as a dependent variable;
Generating a plurality of models to simulate a correlation between the operating factor and the treated water quality; And
And selecting an optimal model based on the prediction accuracy of each model.
12. The method of claim 11,
Wherein the step of generating a plurality of models to simulate a correlation between the operating factor and the quality of the treated water
Inputting the selected independent variable and dependent variable into a multiple regression analysis model;
Performing a plurality of multiple regression analysis in which conditions are set differently;
Calculating standardized regression coefficients of the multiple regression analysis models;
Calculating prediction accuracy of the multiple regression analysis models;
Determining whether the prediction accuracy is within a predetermined allowable range; And
And selecting a multiple regression analysis model having a highest sum of standardized regression coefficients of controllable operating factors of each of the plurality of multiple regression analysis models when the prediction accuracy is within a predetermined allowable range as a result of the determination Processing method.
13. The method of claim 12,
Further comprising the step of removing the independent variable and re-executing the step-selection method multiple regression analysis when a multi-collinearity problem occurs between the independent variables.
13. The method of claim 12,
The step of calculating the standardized regression coefficient
And calculating the normalized regression coefficient when applying the test for the outlier value and the influence value and when the test for the outlier value and the influence value is not used.
12. The method of claim 11,
The prediction accuracy is
A root mean square error (RMSE), and a mean absolute percentage error (MAPE).
12. The method of claim 11,
Wherein the step of generating a plurality of models to simulate a correlation between the operating factor and the quality of the treated water
Inputting the selected independent variable and dependent variable into an artificial neural network model;
Setting parameters for data learning of the artificial neural network model;
Performing data learning of a plurality of artificial neural network models in which the parameters are set differently;
Calculating prediction accuracy of the plurality of artificial neural network models;
Determining whether the prediction accuracy is within a predetermined allowable range; And
And selecting an artificial neural network model having the highest sensitivity of the controllable operation factors of each of the plurality of artificial neural network models when the prediction accuracy is within a preset allowable range as a result of the determination.
17. The method of claim 16,
Wherein the parameter for data learning of the artificial neural network model includes at least one of a number of hidden layers, a threshold, an algorithm to be used, and an error value.
12. The method of claim 11,
Wherein the step of generating a plurality of models to simulate a correlation between the operating factor and the quality of the treated water
Inputting the dependent variable into an ASM (Activated Sludge Models) model;
Determining a reaction formula of a target substance in the reaction tank included in the water treatment process;
Optimizing the plurality of ASM models by optimizing the reaction equation variable for each of a plurality of ASM models having different variables of the reaction formula; And
Calculating prediction accuracy of the plurality of ASM models;
Determining whether the prediction accuracy is within a predetermined allowable range; And
And selecting the ASM model having the highest sensitivity of the controllable operation factors of each of the plurality of ASM models when the prediction accuracy is within a preset allowable range as a result of the determination.
19. A computer program for executing the method for treating water according to any one of claims 10 to 18.

Images