KR20210030311A - Apparatus and method for auto-processing modeling using matlab - Google Patents

Abstract

The present invention relates to a technique for data analysis and optimization which automates correction inputs of parameters and boundary conditions which were relied upon manual work in Matlab and carries out parameter optimization, model uncertainty, and parameter sensitivity analysis. A Matlab modeling automation apparatus comprises: an input unit which receives data used in variable creation; a variable creation unit which uses data received from the input unit to create parameters, and uses model error values generated after executing a Matlab model to optimize parameters; an application unit which applies parameters created by the variable creation unit to an input file of the Matlab model; an execution unit which executes the Matlab model when the application unit applies the parameters to the input file; and a postprocessing unit which uses execution result values of the Matlab model executed by the execution unit to generate model error values.

Description

Modeling automation device and method using MATLAB {APPARATUS AND METHOD FOR AUTO-PROCESSING MODELING USING MATLAB}

The present invention relates to a modeling automation apparatus and method using MATLAB, and more particularly, to automate correction input of boundary conditions and parameters, which had been dependent on conventional manual work in MATLAB, and parameter optimization, model uncertainty. And it relates to a data analysis and optimization technology for automatically performing the sensitivity analysis of the parameters.

MATLAB is an engineering software that provides a numerical analysis and programming environment developed by MathWorks. It supports a matrix-based calculation function, and provides a function or data drawing function and an algorithm implementation through programming. MATLAB is used in various fields of science and engineering that require numerical calculations.

MATLAB can be viewed as a programming language and as an application, but both characteristics may appear depending on the purpose of the person using it. The number of users is as high as 17th among currently used programming languages.

Matlab provides useful toolboxes such as interface with hardware such as several motors and image processing, so it can be easily used in science and engineering departments and engineering fields. For example, if you want to test a visual flight algorithm by installing a camera on a quadcopter with four motors, you need motor control, communication with drones, image processing, position/position estimation, and position/position control algorithms. In other development environments, there is an environment that provides each function, but it is rare except for MATLAB to provide all related libraries.

Referring to Fig. 1, MATLAB provides a desktop environment suitable for the mathematical analysis and design process. In addition, it is possible to apply various algorithms by using an application program.

Referring to FIG. 2, in order to execute the current water-based model, each step of (1) adding a boundary condition, (2) correcting parameters, (3) executing the model, and (4) checking output and drawing a graph is described in EFDC Explorer. You have to manually click one by one of the related icons in the interface called.

Boundary conditions must be frequently modified in order to test models of various scenario conditions. Currently, it is inconvenient, time consuming to modify, and mistakes are frequently made because the icon in the interface must be manually clicked each time the boundary conditions are modified. There was a problem that occurred.

In addition, since parameter correction is performed by manually changing values one by one on the interface, it is labor-intensive, and results are inaccurate because optimization or sensitivity analysis cannot be performed.

Korean Registered Patent Publication No. 10-1670307

Accordingly, the present invention has been proposed to solve the problems of the prior art, and the object of the present invention is to automate the correction input of boundary conditions and parameters that have been relied on conventional manual work in MATLAB, and parameter optimization. In addition, the task is to provide a data analysis and optimization technology that automatically performs model uncertainty and parameter sensitivity analysis.

In order to achieve the above object, a modeling automation apparatus using MATLAB according to the technical idea of the present invention includes: an input unit for receiving data used to generate a variable; A variable generator for generating a parameter using data received from the input unit and optimizing the parameter using a model error value generated after execution of a MATLAB model; An application unit that applies the parameters generated by the variable generation unit to the input file of the MATLAB model; An execution unit that executes the MATLAB model when the application unit applies a parameter to the input file; And a post-processing unit that generates a model error value by using the execution result value of the MATLAB model executed by the execution unit.

In addition, the model may be characterized in that EFDC (Environmental Fluid Dynamics Code).

In addition, the application unit may be characterized in that it applies the parameter to the efdc.inp file.

In addition, the data received by the input unit may include a boundary condition, and the application unit may apply the boundary condition to a qser.inp file.

In addition, the variable generator may be characterized by using a Latin Hypercube-One-factor-At-a-Time (LH-OAT) and a pattern search algorithm.

In addition, the variable generator may be characterized in that it generates a parameter to be applied to the input file of the MATLAB model using LH-OAT.

In addition, the variable generator may be characterized in that it searches for a parameter in which a model error value satisfies a preset criterion using a pattern search algorithm.

In addition, the data received by the input unit may include an observation value, and the post-processing unit may generate a model error value using a result of executing the MATLAB model and the observed value.

In addition, the post-processing unit may analyze the sensitivity of the parameter using the model error value and the Elementary Effect Test (EET).

In addition, the post-processing unit may generate a graph using a result of executing the MATLAB model, a model error, and a sensitivity of a parameter.

On the other hand, in order to achieve the above object, a method for modeling automation using MATLAB according to the technical idea of the present invention includes the steps of: receiving data used for generating a variable by an input unit; Generating a plurality of parameters by using the data received from the input unit by the variable generation unit and selecting one parameter to be applied to the MATLAB model; Applying, by an application unit, the parameter selected by the variable generation unit to the input file of the MATLAB model; Executing the MATLAB model by an execution unit; Generating, by a post-processing unit, a model error value using an execution result value of the MATLAB model executed by the execution unit; Checking, by the variable generation unit, whether the model error value satisfies a preset criterion; And if the model error value does not satisfy a preset criterion, the variable generation unit selects another parameter to be applied to the MATLAB model among a plurality of parameters, and if the model error value satisfies a preset criterion, It is characterized in that the variable generator determines that the parameter applied to the MATLAB model is an optimal parameter.

In addition, the model may be characterized in that EFDC (Environmental Fluid Dynamics Code).

Also, the input file may be an efdc.inp file.

Further, the data includes a boundary condition, and the step of applying the parameter to the input file of the MATLAB model by the application unit may be characterized in that the boundary condition is applied to the qser.inp file.

In addition, the step of generating a plurality of parameters by the variable generation unit using data received from the input unit and selecting one parameter to be applied to the MATLAB model may include: Latin Hypercube-One-factor-At-LH-OAT (Latin Hypercube-One-factor-At- a-Time) and a pattern search algorithm may be used.

In addition, the step of generating a plurality of parameters by the variable generation unit using the data received from the input unit and selecting one parameter to be applied to the MATLAB model may include generating a plurality of parameters using LH-OAT. It can be characterized by that.

In addition, the step of generating a plurality of parameters by the variable generation unit using data received from the input unit and selecting one parameter to be applied to the MATLAB model may include one of the plurality of parameters using a pattern search algorithm. It may be characterized by selecting a parameter.

In addition, the data received by the input unit includes an observation value, and in the post-processing unit generating a model error value using the execution result value of the MATLAB model, the execution result value and the observation value of the MATLAB model are used. Thus, it may be characterized in that the model error value is generated.

In addition, after the post-processing unit generates a model error value using the execution result value of the MATLAB model executed by the execution unit, the post-processing unit uses the model error value and the Elementary Effect Test (EET) to generate a model error value. It may be characterized in that it further comprises the step of analyzing the sensitivity of the applied parameter.

In addition, after the post-processing unit analyzes the sensitivity of the parameter applied to the MATLAB model using the model error value and the Elementary Effect Test (EET), the post-processing unit performs an execution result value of the MATLAB model, a model error value, and It may be characterized in that it further comprises the step of generating a graph using the sensitivity of the array.

According to the modeling automation apparatus and method utilizing MATLAB according to the present invention,

First, the present invention is capable of automating the input process conventionally performed manually, parameter optimization, model uncertainty analysis, parameter sensitivity analysis, graph generation, and error calculation, which were difficult to apply. You will be able to save time.

Second, since the present invention can be applied not only to the Environmental Fluid Dynamics Code (EFDC) model, but also to various models, it is possible to more easily perform a simulation using MATLAB, thereby contributing greatly to research and development.

1 is a diagram showing a GUI of MATLAB executed on a computer device.
FIG. 2 is a diagram showing a process of manually adding a boundary condition and correcting a parameter to query a result value in a conventional MATLAB EFDC (Environmental Fluid Dynamics Code) model.
3 is a block diagram showing the configuration of a modeling automation device using MATLAB according to an embodiment of the present invention.
4 is an exemplary diagram showing a plurality of parameters generated by a variable generation unit using LH-OAT.
5 is an exemplary diagram showing a process of updating a parameter in an input file by a variable generation unit using a pattern search algorithm.
6 is a diagram showing the inquiry contents and parameters of the efdc.inp file in red boxes.
FIG. 7 is a diagram showing a time series data area to which time series data of an inquiry content of a qser.inp file and boundary condition is applied with a red box.
8 is an exemplary diagram showing the result of calculating the sensitivity of a parameter using EET.
9 is a graph showing the sensitivity of parameters calculated using EET.
FIG. 10 is a view showing an example of a result of executing an EFDC model and a result of being mapped to a map using information on the behavior of substances and fluids in a water system.
11 is an exemplary view showing an execution result of a chemical accident simulation model.
12 is a flow chart showing each step of a modeling automation method using MATLAB according to an embodiment of the present invention.

With reference to the accompanying drawings will be described in detail with respect to the modeling automation apparatus and method using MATLAB according to embodiments of the present invention. Since the present invention can apply various changes and have various forms, specific embodiments will be illustrated in the drawings and described in detail in the text. However, this is not intended to limit the present invention to a specific form of disclosure, it should be understood to include all changes, equivalents, and substitutes included in the spirit and scope of the present invention. In describing each drawing, similar reference numerals have been used for similar elements.

In addition, unless otherwise defined, all terms used herein including technical or scientific terms have the same meaning as commonly understood by one of ordinary skill in the art to which the present invention belongs. Terms such as those defined in commonly used dictionaries should be interpreted as having a meaning consistent with the meaning in the context of the related technology, and should not be interpreted as an ideal or excessively formal meaning unless explicitly defined in the present application. Does not.

The modeling automation device and method using MATLAB according to an embodiment of the present invention is installed and operated on a single computer device on which MATLAB is installed, or is distributedly installed on a plurality of computer devices, and each computer device is connected to each other by wire or wirelessly. It can be implemented as being interlocked.

MATLAB provides an environment in which various simulation models can be used. For example, models available in MATLAB include EFDC (Environmental Fluid Dynamics Code) and chemical accident simulation.

Referring to FIG. 3, the MATLAB modeling automation apparatus according to an embodiment of the present invention generates a parameter using data received from an input unit 110 and an input unit 110 for receiving data used for variable generation, and (MATLAB) A variable generation unit 120 that optimizes the parameters using model error values generated after the model is executed, and an application unit that applies the parameters generated by the variable generation unit 120 to the input file of the MATLAB model. (140), When the application unit 140 applies the parameter to the input file, the execution unit 160 that executes the MATLAB model, and the model error using the execution result value of the MATLAB model executed by the execution unit 160 It includes a post-processing unit 180 for generating a value.

In the following, as an embodiment, a description will be made on the basis that the MATLAB model is EFDC (Environmental Fluid Dynamics Code).

The EFDC water system model is a functional surface water modeling system that includes fluid dynamics, sediment contamination and eutrophication components. EFDC is a hydrodynamic model that can be used to simulate hydraulic systems in one, two and three dimensions. EFDC is used in a variety of water environments including rivers, lakes, reservoirs, wetlands, estuaries and coastal waters to support environmental assessment, management and regulatory requirements.

EFDC uses extended or sigma vertical coordinates, orthogonal, curved or rectangular horizontal coordinates to represent the physical properties of a water body. EFDC can calculate the equations of three-dimensional vertical constant, free surface, and turbulent mean motion for variable density fluids. In addition, dynamically coupled transport equations for turbulent kinetic energy, turbulent length scale, salinity and temperature can be solved.

In Korea, the EFDC model is frequently used by the National Institute of Environmental Sciences, its affiliated institutions, and research institutes related to environmental engineering.

MATLAB supports the simulation of EFDC models. In MATLAB, the files related to the EFDC model include efdc.inp and qser.inp.

The variable generator 120 uses LH-OAT (Latin Hypercube-One-factor-At-a-Time) and pattern search algorithms to generate and optimize parameters.

The LH-OAT method is the Latin Hypercube (LH) method, which analyzes the global sensitivity using a multivariate linear regression method after dividing the parameter space into N intervals, randomly extracting the variables for each interval, and parameters. It is a technique that combines the One-Factor-At-a-Time (OAT) method, which analyzes local sensitivity by changing parameters one by one while other parameters are fixed in space. The LH-OAT method can effectively perform calibration and selection of parameters for sensitivity analysis.

The variable generator 120 generates a plurality of parameters to be applied to the input file of the MATLAB model using LH-OAT. Referring to FIG. 4, the variable generation unit 120 generates a plurality of parameters OAT using LH-OAT. One parameter consists of multiple values to form a set. At this time, the generated parameter array is individually applied to the EFDC model and used for model execution.

The pattern search algorithm is one of the numerical optimization methods. The pattern search algorithm performs optimization by checking the error by changing the parameter value in the multidimensional analysis space, reducing the step size when the error is reduced, and searching for a solution with the least error.

The variable generator 120 searches for a parameter whose model error value satisfies a preset criterion among a plurality of parameters previously generated using LH-OAT using a pattern search algorithm. Referring to FIG. 5, the variable generator 120 selects one parameter to be applied to the MATLAB model from among a plurality of parameters. When the selected parameter is applied and the error value of the executed MATLAB model does not satisfy a preset criterion, the variable generator 120 selects another parameter from among a plurality of parameters. The other parameters selected by the variable generation unit 120 are again applied to the MATLAB model and executed. The variable generator 120 searches for an optimal parameter by applying a plurality of parameters to the MATLAB model one by one and comparing the results.

It is preferable that the predetermined criterion of the model error value is a size sufficient to allow errors. For example, among the model error indicators, NSE (Nash Sutcliffe Efficiency) has a value between -∞ and 1, and the NSE value has a value of 1 when there is no error in the model value, and -∞ when the error is maximum. In general, if the NSE value is greater than 0.5, it is determined that the model performance is sufficient, so 0.5 can be set as a preset criterion for determining the model error. At this time, if the NSE value of the executed MATLAB model is less than 0.5, the variable generator 120 selects another parameter to be applied to the MATLAB model from among the plurality of parameters. If the model's NSE value is greater than 0.5, the parameter applied to the MATLAB model satisfies the preset criterion. Therefore, it is determined that the parameter is an optimized parameter, and the process of selecting another parameter is terminated.

As another embodiment, the preset criterion may be a parameter having the smallest model error value among all parameters generated by the variable generator 120. In this case, a plurality of parameters are applied and executed one by one to the MATLAB model to search for the case with the smallest model error value.

The application unit 140 applies the parameter selected by the variable generation unit 120 to the efdc.inp file.

Conventionally, efdc.inp or qser.inp is updated when parameters or boundary condition data are manually modified in the EFDC interface (see FIG. 2 (1)). Updates of efdc.inp and qser.inp are possible using the EFDC interface, but users can modify them artificially by directly accessing the path where the efdc.inp and qser.inp files are stored and executing the corresponding file with an editor.

The application unit 140 automatically searches the efdc.inp file and applies the parameter value to the efdc.inp file. At this time, the application includes additions, substitutions, and deletions. 6 shows some of the inquiry contents of the efdc.inp file, and the red box area is a parameter. In addition, information on the type of boundary condition is labeled in the efdc.inp file. The application unit 140 applies the value of the parameter selected by the variable generation unit 120 to an area related to the parameter as shown in the red box of FIG. 6.

In addition, this embodiment further includes a boundary condition in the data received by the input unit 110. Boundary conditions are conditions that are additionally limited when calculating differential equations in numerical analysis models. For example, boundary conditions of a water system model such as EFDC may be upstream and downstream water levels, flow rates, and water quality. Boundary conditions take the form of time series data. 7 shows a part of the inquiry contents of the qser.inp file, and the red box area is time series data of the boundary condition. The application unit 140 applies the boundary condition to the qser.inp file by correcting the existing time series data recorded in the qser.inp file as the boundary condition time series data. In addition, the application unit 140 corrects the information labeled in the efdc.inp file and the time series data in the qser.inp file together if the type of the boundary condition is to be corrected.

When the application unit 140 applies the parameters and boundary conditions to the input file of the MATLAB model, the execution unit 160 executes the MATLAB model.

After the MATLAB model is executed, when the Matlab operation is finished, the post-processing unit 180 collects the execution result value from MATLAB. The data received by the input unit 110 of this embodiment further includes an observation value. In order to calculate the error of the model, a true value as a reference is required. In the water system model, actual observed data are used as true values. Actually observed data include water level, water temperature, salinity, and chemical concentration. These actual observed data become observed values.

The post-processing unit 180 generates a model error value by using the result value and the observed value of the MATLAB model. The model error can be calculated using NSE and Root Mean Square Deviation (RMSD).

The post-processing unit 180 analyzes the sensitivity of the parameter applied to the current MATLAB model by using the model error value and the Elementary Effect Test (EET). 8 is an example showing a result of analyzing the sensitivity of each value of a parameter having five values.

The Elementary Effect Test (EET) method is a global sensitivity analysis technique, and is used to analyze sensitivity in models with a large number of input data or models with complex calculation processes.

The mean(EE) value appears larger as the parameter value has a greater influence on the model's execution result. In addition, the greater the difference in the effect of the current parameter on the model according to the combination with other parameters, the larger the std(EE) value appears.

A method in which the post-processing unit 180 calculates the sensitivity of the parameter using EET is as shown in Equation 1.

[Equation 1]

=a,

=b,

=c가 된다. 이 중 민감도를 분석하려는 매개변수가 b인 경우 i는 2가 된다. 매개변수가 총 3개이므로, k=3이 된다. b=2일 때 모델을 실행한 결과와, b=3일 때 모델을 실행한 결과를 비교하면 b에 대한 민감도를 분석할 수 있고, 이때의 Δ는 1이 된다(2→3).In this case, the EE (elementary effect) is an index showing sensitivity. X is the parameter applied to the model. Y is the value of the model. Since the model error value is a value obtained by subtracting the true value from the execution result value, both the execution result value and the model error value can be used for the model value. i is the index of the parameter to be analyzed for sensitivity, k is the total number of parameters applied to the model, and Δ is the increment of the parameter value changed for the sensitivity analysis. For example, assuming that the model has parameters a, b, c applied,

=a,

=b,

=c. Among these, if the parameter to be analyzed for sensitivity is b, i becomes 2. Since there are a total of 3 parameters, k=3. By comparing the result of executing the model when b=2 and the result of executing the model when b=3, the sensitivity to b can be analyzed, and Δ at this time becomes 1 (2→3).

9 is an exemplary graph showing the result of analyzing the sensitivity of a parameter using EET. In this example graph, it is shown that the sensitivity of the AHO factor is the highest among the parameters.

The post-processing unit 180 generates a graph using the execution result of the MATLAB model, model error, and sensitivity of parameters.

FIG. 10 is a diagram illustrating an EFDC model being executed in MATLAB, obtaining a result value of a substance and fluid behavior in a water system, and mapping it to a map using the result.

11 is a graph showing the change in the concentration of toluene in the water system for each accident by automatically correcting the boundary conditions to fire, leakage, and explosion scenarios in Matlab's chemical accident simulation model as another embodiment.

Next, a method for automating MATLAB modeling according to an embodiment of the present invention will be described.

Referring to FIG. 12, in this embodiment, the input unit 110 receives data used for variable generation (S110), and the variable generation unit 120 uses the data received from the input unit 110 to obtain a plurality of parameters. Steps of creating a variable and selecting one parameter to be applied to the MATLAB model (S122, S126), applying the parameter selected in the variable generation unit 120 by the application unit 140 to the input file of the MATLAB model ( S140), the step of executing the MATLAB model by the execution unit 160 (S160), the step of generating a model error value by the post-processing unit 180 using the execution result value of the MATLAB model executed by the execution unit 160 ( S184), the variable generation unit 120 includes a step (S192) of checking whether the model error value satisfies a preset criterion.

If the model error does not satisfy the preset criterion, the variable generation unit 120 selects another parameter to be applied to the MATLAB model among the plurality of parameters (S126). If the variable generator 120 selects another parameter, it is sequentially executed from step S140 again.

If the model error satisfies a preset criterion, the variable generator 120 determines that the parameter applied to the MATLAB model is an optimal parameter (S194).

If the MATLAB model is EFDC, the input files include the efdc.inp file and the qser.inp file.

In addition, the data received in step S110 includes a boundary condition. At this time, in step S140, the boundary condition is applied to the qser.inp file.

The step of the variable generation unit 120 to generate a plurality of parameters using the data received from the input unit 110 and to select one parameter to be applied to the MATLAB model includes a plurality of parameters using LH-OAT. It consists of a step of generating a variable (S122), and a step (S126) of selecting one of the plurality of parameters using a pattern search algorithm. In step S126, when the model error value does not satisfy a preset criterion, another parameter is selected among a plurality of parameters.

In addition, the data received in step S110 includes the observed value. In this case, in step S184, a model error value is generated by using the result of executing the MATLAB model and the observed value.

In addition, this embodiment further includes a step (S186) of analyzing the sensitivity of the parameter applied to the MATLAB model by the post-processing unit 180 using the model error value and EET after step S184.

In addition, this embodiment further includes the step of generating, by the post-processing unit 180 after step S186, a graph using the execution result value of the MATLAB model, the model error value, and the sensitivity of the array.

Although a preferred embodiment of the present invention has been described above, the present invention can use various changes, modifications, and equivalents. It is clear that the present invention can be applied in the same manner by appropriately modifying the above embodiments. Therefore, the above description is not intended to limit the scope of the present invention determined by the limits of the following claims.

10: MATLAB 100: MATLAB modeling automation device
110: input unit 120: variable generation unit
140: application unit 160: execution unit
180: post-processing unit

Claims (20)

An input unit for receiving data used to generate a variable;
A variable generator for generating a parameter using data received from the input unit and optimizing the parameter using a model error value generated after execution of a MATLAB model;
An application unit that applies the parameters generated by the variable generation unit to the input file of the MATLAB model;
An execution unit that executes the MATLAB model when the application unit applies a parameter to the input file; And
And a post-processing unit that generates a model error value by using an execution result value of the MATLAB model executed by the execution unit. The method of claim 1,
Matlab modeling automation device, characterized in that the model is EFDC (Environmental Fluid Dynamics Code). The method of claim 2,
MATLAB modeling automation device, characterized in that the application unit applies the parameter to the efdc.inp file. The method of claim 3,
The data received by the input unit includes a boundary condition,
The application unit applies the boundary condition to a qser.inp file. The method of claim 2,
The variable generation unit LH-OAT (Latin Hypercube-One-factor-At-a-Time) and a pattern search (Pattern search) algorithm, characterized in that using a MATLAB modeling automation device. The method of claim 5,
The variable generation unit generates a parameter to be applied to the input file of the MATLAB model using LH-OAT. The method of claim 5,
The variable generation unit searches for a parameter in which a model error value satisfies a preset criterion using a pattern search algorithm. The method of claim 5,
The data received by the input unit includes an observation value,
The post-processing unit generates a model error value using a result of executing the MATLAB model and the observed value. The method of claim 5,
The post-processing unit analyzes the sensitivity of the parameter using the model error value and the Elementary Effect Test (EET). The method of claim 9,
The post-processing unit generates a graph using a result of executing the MATLAB model, a model error, and a sensitivity of a parameter. Receiving data used for generating a variable by an input unit;
Generating a plurality of parameters by using the data received from the input unit by the variable generator and selecting one parameter to be applied to the MATLAB model;
Applying, by an application unit, the parameter selected by the variable generation unit to the input file of the MATLAB model;
Executing the MATLAB model by an execution unit;
Generating, by a post-processing unit, a model error value using an execution result value of the MATLAB model executed by the execution unit;
Checking, by the variable generator, whether the model error value satisfies a preset criterion; And
If the model error value does not satisfy a preset criterion, the variable generation unit comprises the step of selecting another parameter to be applied to the MATLAB model from among the plurality of parameters,
When the model error value satisfies a preset criterion, the variable generator determines that the parameter applied to the MATLAB model is an optimal parameter. The method of claim 11,
The matlab modeling automation method, characterized in that the model is EFDC (Environmental Fluid Dynamics Code). The method of claim 12,
The input file is a MATLAB modeling automation method, characterized in that the efdc.inp file. The method of claim 13,
The data includes a boundary condition,
In the step of applying, by the application unit, the parameter to the input file of the MATLAB model, the boundary condition is applied to a qser.inp file. The method of claim 12, wherein the variable generation unit generates a plurality of parameters using data received from the input unit and selects one parameter to be applied to the MATLAB model,
A MATLAB modeling automation method characterized by using a LH-OAT (Latin Hypercube-One-factor-At-a-Time) and a pattern search algorithm. The method of claim 15, wherein the variable generation unit generates a plurality of parameters using data received from the input unit and selects one parameter to be applied to the MATLAB model,
MATLAB modeling automation method, characterized in that generating a plurality of parameters using LH-OAT. The method of claim 15, wherein the variable generation unit generates a plurality of parameters using data received from the input unit and selects one parameter to be applied to the MATLAB model,
MATLAB modeling automation method, characterized in that selecting one of a plurality of parameters using a pattern search algorithm. The method of claim 14,
The data received by the input unit includes an observation value,
The step of generating, by the post-processing unit, a model error value using the execution result value of the MATLAB model, comprises generating a model error value using the execution result value and the observation value of the MATLAB model. . The method of claim 14,
After the post-processing unit generates a model error value using the execution result value of the MATLAB model executed by the execution unit,
And analyzing, by the post-processing unit, the sensitivity of the parameter applied to the MATLAB model using the model error value and Elementary Effect Test (EET). The method of claim 19, wherein after the post-processing unit analyzes the sensitivity of the parameter applied to the MATLAB model using the model error value and the Elementary Effect Test (EET),
And generating, by the post-processing unit, a graph using an execution result value of the MATLAB model, a model error value, and a sensitivity of an array.

Images