# FLOW ANALYSIS APPARATUS AND METHOD THEREFOR

A flow analysis apparatus is provided. The flow analysis apparatus includes a model deriver configured to generate a flow analytic model for performing a flow analysis for a plurality of cells by using analytic data including a plurality of input signals used for performing multiple times iterations of numerical analysis by Computational Fluid Dynamics (CFD) and a plurality of output signals corresponding to each of the plurality of input signals, and a flow analyzer configured to perform the flow analysis for the plurality of cells that divide the space around a design target component by using the generated flow analytic model.

**Description**

**CROSS-REFERENCE TO RELATED APPLICATION**

This application claims priority to Korean Patent Application No. 10-2018-0097538, filed on Aug. 21, 2018, the disclosure of which is incorporated by reference herein in its entirety.

**BACKGROUND**

**Field**

Apparatuses and methods consistent with exemplary embodiments relate to a flow analysis technology, and more particularly, to a flow analysis apparatus and a method therefor.

**Description of the Related Art**

Flow analysis means to confirm an interaction between fluids such as liquids and gases around a target component to be analyzed and a surface defined by a boundary condition, and a change in flow thereby and relevant characteristics. Computational Fluid Dynamics (CFD) is to reproduce the flow of heat and fluid through computational operation, and has reproduced the analysis for the past heat and fluid motion by effective numerical analysis for a short time due to the development of a computer rather than a method based on the experiment, thereby resulting in time and cost savings. A governing equation of fluid behavior is a nonlinear partial differential equation describing the motion of fluid with viscosity, which is an equation including both the convection term and the diffusion term, and can analyze most flows that are present in the natural world such as weather and the fluid flow of current.

**SUMMARY**

Aspects of one or more exemplary embodiments provide a flow analysis apparatus and a method therefor for shortening the time of flow analysis.

According to an aspect of an exemplary embodiment, there is provided a flow analysis apparatus including: a model deriver configured to generate a flow analytic model for performing a flow analysis for a plurality of cells by simulating a numerical analysis by Computational Fluid Dynamics (CFD) with respect to the plurality of cells that divide a space around a component, and a flow analyzer configured to perform the flow analysis for the plurality of cells that divide the space around a design target component by using the generated flow analytic model.

The model deriver may include an analyzing data storage configured to store analytic data used for the numerical analysis, a signal generating model deriver configured to generate a signal generating model for predicting an input signal contributing to an output signal of the numerical analysis performed multiple times iterations through the analytic data, and an analytic model deriver configured to generate the analytic model for predicting the output signal of the numerical analysis performed multiple times iterations through the analytic data.

The signal generating model deriver predicts the input signal contributing to the output signal of the numerical analysis performed multiple times iterations through an Equation

where the k and the K refer to the number of times of numerical analysis, the H refers to a degree of influence, the Q refers to a weight, the D refers to a cancellation constant, the V(k) refers to an input signal of the k^{th }numerical analysis, the Y(k) refers to an output signal of the k^{th }numerical analysis, and the {circumflex over (V)}(k+T) refers to an input signal of the (k+T)^{th }numerical analysis.

The signal generating model deriver generates a signal generating model by deriving a parameter through an optimization algorithm after constituting a relationship equation of the signal generating model where the parameter is not determined.

The analytic model deriver predicts the output signal of the numerical analysis performed multiple times iterations through an Equation

where the k and the T refer to parameters representing the number of times of numerical analysis, the A refers to a degree of influence, the P refers to a weight, the C refers to a cancellation constant, the Y(k) refers to an output signal of the k^{th }numerical analysis, the {circumflex over (V)}(k+T) refers to an input signal of the (k+T)^{th }numerical analysis, and the Ŷ(k+T) refers to an output signal of the (k+T)^{th }numerical analysis.

The analytic model deriver generates an analytic model by deriving a parameter through an optimization algorithm after constituting a relationship equation of the analytic model where the parameter is not determined.

The flow analyzer may include a numerical analyzer configured to derive the analytic data by performing the numerical analysis with respect to a design target component, a signal generator configured to generate the input signal contributing to the output signal of the numerical analysis performed multiple times iterations through the signal generating model derived from the signal generating model deriver and the analytic data derived from the numerical analyzer, and an analyzer configured to derive the output signal of the numerical analysis performed multiple times iterations through the analytic model derived from the analytic model deriver and the input signal predicted by the signal generator.

The numerical analyzer derives the analytic data by performing the numerical analysis by Computational Fluid Dynamics with respect to the plurality of cells that divide the space around a design target component.

The signal generator derives the input signal contributing to the output signal of the numerical analysis performed the number of predetermined times iterations by inputting the analytic data to the signal generating model.

The analyzer derives the output signal of the numerical analysis where the numerical analysis has been performed the number of predetermined times iterations by inputting the output signal of the analytic data and the contributing input signal to the analytic model.

According to an aspect of another exemplary embodiment, there is provided a flow analysis apparatus including: a model deriver configured to generate a flow analytic model for performing a flow analysis for a plurality of cells by using analytic data used for a numerical analysis by Computational Fluid Dynamics (CFD) with respect to the plurality of cells that divide a space around a component, and a flow analyzer configured to perform the flow analysis for the plurality of cells that divide the space around a design target component by using the generated flow analytic model.

The flow analytic model may simulate the numerical analysis performed multiple times iterations by the Computational Fluid Dynamics.

The flow analytic model may include one or more signal generating model for predicting an input signal contributing to a result of the numerical analysis performed multiple times iterations among the plurality of input signals through the analytic data, and an analytic model for predicting the result of the numerical analysis performed multiple times iterations through the analytic data.

The signal generating model may predict the input signal contributing to the output signal of the numerical analysis performed multiple times iterations through an Equation

where the k and T refer to the number of times of numerical analysis, the H refers to a degree of influence, the Q refers to a weight, the D refers to a cancellation constant, the V(k) refers to an input signal of the k^{th }numerical analysis, the Y(k) refers to an output signal of the k^{th }numerical analysis, and the {circumflex over (V)}(k+T) refers to an input signal of the (k+T)^{th }numerical analysis.

The analytic model may predict the output signal of the numerical analysis performed multiple times iterations through an Equation

where the k and the T refer to the number of times of numerical analysis, the A refers to a degree of influence, the P refers to a weight, the C refers to a cancellation constant, the Y(k) refers to an output signal of the k^{th }numerical analysis, the {circumflex over (V)}(k+T) refers to an input signal of the (k+T)^{th }numerical analysis, and the Ŷ(k+T) refers to an output signal of the (k+T)^{th }numerical analysis.

According to an aspect of another exemplary embodiment, there is provided a flow analysis method including: storing, by a model deriver, analytic data including a plurality of input signals used for a numerical analysis by Computational Fluid Dynamics (CFD) with respect to a plurality of cells that divide a space around a component and a plurality of output signals corresponding to each of the plurality of input signals, generating, by the model deriver, a flow analytic model for performing a flow analysis for the plurality of cells by using the analytic data, and performing, by a flow analyzer, the flow analysis for the plurality of cells that divide the space around a design target component by using the flow analytic model.

The generating the flow analytic model includes generating, by a signal generating model deriver, a signal generating model for predicting an input signal contributing to an output signal of the numerical analysis performed multiple times iterations among the plurality of input signals through the analytic data, and generating, by an analytic model deriver, an analytic model for predicting the output signal of the numerical analysis performed multiple times iterations among the plurality of input signals through the analytic data.

The generating the signal generating model includes predicting the input signal contributing to the output signal of the numerical analysis performed multiple times iterations through an Equation

where the k and the T refer to the number of times of numerical analysis, the H refers to a degree of influence, the Q refers to a weight, the D refers to a cancellation constant, the V(k) refers to an input signal of the k^{th }numerical analysis, the Y(k) refers to an output signal of the k^{th }numerical analysis, and the {circumflex over (V)}(k) refers to an input signal of the (k+T)^{th }numerical analysis.

The generating the analytic model includes predicting the output signal of the numerical analysis performed multiple times iterations through an Equation

where the k and the T refer to the number of times of numerical analysis, the A refers to a degree of influence, the P refers to a weight, the C refers to a cancellation constant, the Y(k) refers to an output signal of the k^{th }numerical analysis, the {circumflex over (V)}(k+T) refers to an input signal of the (k+T)^{th }numerical analysis, and the Ŷ(k+T) refers to an output signal of the (k+T)^{th }numerical analysis.

The generating the signal generating model includes constituting, by the signal generating model deriver, a relationship equation of the signal generating model where a parameter is not determined, and completing, by the signal generating model deriver, the signal generating model by deriving the parameter through an optimization algorithm.

The generating the analytic model includes constituting, by the analytic model deriver, a relationship equation of the analytic model where a parameter is not determined, and completing, by the analytic model deriver, the analytic model by deriving the parameter through an optimization algorithm.

The performing the flow analysis may include deriving, by a numerical analyzer, the analytic data by performing the numerical analysis with respect to a design target component, deriving, by a signal generator, the input signal contributing to the output signal of the numerical analysis performed multiple times iterations through the signal generating model derived from the signal generating model deriver and the analytic data derived from the numerical analyzer, and deriving, by an analyzer, the output signal of the numerical analysis performed multiple times iterations through the analytic model derived from the analytic model deriver, the input signal derived from the signal generator and the analytic data derived from the numerical analyzer.

As described above, according to one or more exemplary embodiments, it is possible to shorten the time for performing the flow analysis, thereby shortening the time required to develop the component.

**BRIEF DESCRIPTION OF THE DRAWINGS**

The above and other aspects will become more apparent from the following description of the exemplary embodiments with reference to the accompanying drawings, in which:

**DETAILED DESCRIPTION**

Hereinafter, various modifications and various embodiments will be described in detail with reference to the accompanying drawings so that those skilled in the art can easily carry out the disclosure. It should be understood, however, that the various embodiments are not for limiting the scope of the disclosure to the specific embodiment, but they should be interpreted to include all modifications, equivalents, and alternatives of the embodiments included within the spirit and technical scope disclosed herein. In order to clearly illustrate the disclosure in the drawings, some of the elements that are not essential to the complete understanding of the disclosure may be omitted, and like reference numerals refer to like elements throughout the specification.

The terminology used in the disclosure is for the purpose of describing specific embodiments only and is not intended to limit the scope of the disclosure. The singular expressions “a”, “an”, and “the” are intended to include the plural expressions as well unless the context clearly indicates otherwise. In the present disclosure, terms such as “comprises,” “includes,’ or “have/has” should be construed as designating that there are such features, integers, steps, operations, components, parts, and/or combinations thereof, not to exclude the presence or possibility of adding of one or more of other features, integers, steps, operations, components, parts, and/or combinations thereof.

Further, terms such as “first,” “second,” and so on may be used to describe a variety of elements, but the elements should not be limited by these terms. The terms are used simply to distinguish one element from other elements. The use of such ordinal numbers should not be construed as limiting the meaning of the term. For example, the components associated with such an ordinal number should not be limited in the order of use, placement order, or the like. If necessary, each ordinal number may be used interchangeably.

First, a flow analytic model according to an exemplary embodiment will be described.

Referring to

Referring to

Referring to a graph of

Therefore, according to one or more exemplary embodiments, the result of the numerical analysis performed multiple times iterations by Computational Fluid Dynamics may be obtained by using the analytic data including a plurality of input signals used for the numerical analysis and a plurality of output signals corresponding to the plurality of input signals. That is, a flow analytic model for predicting the output signal is generated, and the flow analysis is performed by using the generated flow analytic model. Therefore, it is possible to reduce the time for obtaining the approximate solution of the partial differential equation, thereby shortening the flow analysis time. Therefore, it is possible to shorten the time for designing the component.

The flow analytic model can use at least one of the models listed Table 1.

In addition, the flow analytic model can be derived by using at least one of the optimization algorithms listed in Table 2.

Next, a flow analysis apparatus according to an exemplary embodiment will be described. **10** may include a model deriver **100** and a flow analyzer **200**.

The model deriver **100** generates a flow analytic model for performing a flow analysis for a plurality of cells that divide a space around a target. At this time, the model deriver **100** generates a flow analytic model by using the analytic data used for the numerical analysis by Computational Fluid Dynamics (CFD). Herein, the analytic data includes a plurality of input signals used for the numerical analysis performed multiple times iterations and a plurality of output signals corresponding to the plurality of input signals. The flow analytic model derives a result of the numerical analysis performed iterations by Computational Fluid Dynamics by simulating the numerical analysis by Computational Fluid Dynamics.

In particular, the flow analytic model can be composed of a plurality of models. The flow analytic model can include one or more signal generating models and one or more analytic models. The model deriver **100** includes an analytic data storage **110**, a signal generating model deriver **120**, and an analytic model deriver **130**. The signal generating model and the analytic model can also use any one of the plurality of models of Table 1.

The analytic data storage **110** stores the analytic data. The analytic data can become analytic data used for the numerical analysis by Computational Fluid Dynamics for the plurality of cells CE that deride the area around the component CP. This analytic data includes a plurality of input signals and a plurality of output signals corresponding to the plurality of input signals. For example, the input signal can be the laminar flow viscosity of the fluid, the turbulent conduction, the time difference between the numerical analyses performed iterations, etc. in each cell CE. The output signal may be the characteristics of the fluid. For example, the output signal can be the density, the momentum in the x and y directions, the internal energy, etc. in each cell CE.

The signal generating model deriver **120** generates a signal generating model for deriving the input signal contributing to the output signal of the numerical analysis performed multiple times iterations among the plurality of input signals by using the analytic data stored in the analytic data storage **110**. For example, if there are a plurality of input signals, the input signal for determining the output signal of the numerical analysis after performing the numerical analysis multiple times iterations can become a part of the plurality of input signals. As described above, the input signal contributing to the output signal means an input signal of the type and the iteration timing, which affect a change in the value of the output signal, among the input signals of a plurality of types and the iteration timings. That is, the signal generating model is for predicting the input signal contributing to the output signal after performing the numerical analysis multiple times iterations.

According to an exemplary embodiment, the signal generating model is as in Equation 1.

In Equation 1, k and T refer to the number of times of numerical analysis. V(k) refers to the input signal of the k^{th }numerical analysis, Y(k) refers to the output signal of the k^{th }numerical analysis, and {circumflex over (V)}(k+T) refers to the input signal of the (k+T)^{th }numerical analysis. Herein, T can have a different value or the same value according to the type of the output signal to be predicted (e.g., a density, a momentum in the x and y directions, an internal energy, etc.). H refers to a degree of influence on each cell. For example, only values corresponding to the corresponding cell affect the prediction of a value of any one cell, and values of other cells do not affect it. That is, it is a value for selecting a cell influenced by the corresponding input signal or output signal. Q refers to a weight. That is, the weight Q means a degree influenced by V(k) and Y(k) on the output {circumflex over (V)}(k+T). D refers to a cancellation constant for canceling a modeling error. That is, referring to the Equation 1 and the graph of ^{th }numerical analysis, that is, the input signal {circumflex over (V)}(k+T) contributing to the output signal according to the (k+T)^{th }numerical analysis from the input signal V(k) and the output signal Y(k).

To derive the signal generating model, the signal generating model deriver **120** constitutes a relationship equation where, for example, the degree of influence H, the weight Q, and the cancellation constant D, which are the parameters of the Equation 1, are not determined. The signal generating model deriver **120** derives the parameters, that is, the degree of influence H, the weight Q, and the cancellation constant D through an optimization algorithm by inputting a plurality of analytic data, for example, V(k), Y(k), and V(k+T). The optimization algorithm can be, for example, a Least-Squares Method, a Backpropagation Algorithm, etc. As described above, when the parameters H, Q, and D of the relationship equation are determined, the signal generating model as in Equation 1 is completed.

For example, the signal generating model deriver **120** can constitute an Artificial Neural Network having a transfer function where the weight connections as in the Equation 1 are the parameters H, Q, and D. The parameters H, Q, and D, which are the weight connections, can be derived through the optimization algorithm (e.g., the Backpropagation algorithm) by using the analytic data as learning data, and the artificial neural network where the parameters H, Q and D have been determined can be derived as a signal generating model.

The analytic model deriver **130** derives the analytic model for calculating the output signal of the numerical analysis performed multiple times iterations by using the analytic data that include a plurality of input signals used for the numerical analysis based on the Computational Fluid Dynamics (CFD) and a plurality of output signals corresponding to the plurality of input signals. This analytic model simulates the numerical analysis performed multiple times iterations based on the Computational Fluid Dynamics (CFD).

For example, the analytic model is as in Equation 2.

In the Equation 2, k and T refer to the number of times of numerical analysis, Y(k) refers to the output signal of the k^{th }numerical analysis, {circumflex over (V)}(k+T) refers to the input signal of the (k+T)^{th }numerical analysis, and Ŷ(k+T) refers to the output signal of the (k+T)^{th }numerical analysis. Herein, T can have a different value or the same value according to the type of the output signal to be predicted (e.g., a density, a momentum in the x and y directions, an internal energy, etc.). A refers to a degree of influence on each cell. For example, only values corresponding to the corresponding cell affect the prediction of a value of any one cell and values of other cells do not affect it. That is, it is a value for selecting a cell influenced by the corresponding input signal or output signal. P refers to a weight. That is, the weight P means the degree influenced by Y(k) and {circumflex over (V)}(k+T) on the output Ŷ(k+T). C refers to a cancellation constant for canceling a modeling error.

Referring to the Equation 1, the Equation 2, and the graph of ^{th }numerical analysis from the input signal V(k) and the output signal Y(k) of the k^{th }numerical analysis. The analytic model of the Equation 2 can derive the output signal Y(k) used for the k^{th }numerical analysis, and the output signal Ŷ(k+T) according to the (k+T)^{th }numerical analysis from the input signal {circumflex over (V)}(k+T) of the (k+T)^{th }numerical analysis derived by the signal generating model of the Equation 1.

To derive the analytical model, the analytic model deriver **130** constitutes a relationship equation where, for example, the degree of influence A, the weight P, and the cancellation constant C, which are the parameters of the Equation 2, are not determined. The analytic model deriver **130** derives the parameters, that is, the degree of influence A, the weight P, and the cancellation constant C through an optimization algorithm by inputting a plurality of analytic data, for example, Y(k), V(k+T), and Y(k+T) to the relationship equation where the parameters are not determined. The optimization algorithm can be, for example, a Least-Squares Method, a Backpropagation Algorithm, etc. As described above, when the parameters A, P, and C are determined, the analytical model as in the Equation 2 is completed.

For example, the analytic model deriver **130** can constitute an artificial neural network having a transfer function where the weight connections as in the Equation 2 are the parameters A, P, and C. The parameters A, P, and C, which are the weight connections, can be derived through the optimization algorithm (e.g., the Backpropagation algorithm) by using the analytic data as learning data, and the artificial neural network where the parameters A, P, and C, which are the weight connections, have been determined can be derived by the analytic model.

The flow analyzer **200** performs the flow analysis for the plurality of cells CE that divide the space around the design target component CP by using the flow analytic model including the signal generating model and the analytical model derived from the model deriver **100**. The flow analyzer **200** may include a numerical analyzer **210**, a signal generator **220**, and an analyzer **230**.

The numerical analyzer **210** performs the numerical analysis by Computational Fluid Dynamics (CFD). Therefore, an input signal for the numerical analysis and an output signal corresponding to the input signal are derived. For example, the input signal according to the numerical analysis can be V(k), and the output signal can be Y(k).

The signal generator **220** predicts the input signal contributing to the output signal of the numerical analysis performed multiple times (k+T) iterations by reflecting the input signal and the output signal derived from the numerical analyzer **210** to the signal generating model generated by the signal generating model deriver **120**. For example, the signal generator **220** can derive the input signal {circumflex over (V)}(k+T) contributing to the output signal of the numerical analysis performed multiple times (k+T) iterations by inputting the input signal V(k) and the output signal Y(k) to the signal generating model as in the Equation 1.

The analyzer **230** predicts the output signal of the numerical analysis performed multiple times (k+T) iterations by reflecting the input signal predicted by the signal generator **220** and the output signal derived from the numerical analyzer **210** to the analytical model derived from the analytic model deriver **130**. For example, the analyzer **230** can derive the output signal Ŷ(k+T) of the numerical analysis performed multiple times (k+T) iterations by inputting the predicted input signal {circumflex over (V)}(k+T) and output signal Y(k) to the analytical model as in the Equation 2.

Referring to ^{th }numerical analysis by the numerical analyzer **210**, it is not necessary to perform the numerical analysis the number of times T iterations, thereby shortening the time required for the flow analysis by the time for performing the numerical analysis the number of times T iterations. Therefore, it is possible to shorten the time required for developing the component.

Next, a flow analysis method according to an exemplary embodiment will be described.

Referring to **100** generates the flow analytic model for performing the flow analysis for the plurality of cells CE that divide the space around the flow analysis target component CP by using the analytic data (operation S**110**). Herein, the analytic data includes a plurality of input signals used for the numerical analysis performed multiple times iterations by Computational Fluid Dynamics (CFD) and a plurality of output signals corresponding to the plurality of input signals. In particular, the flow analytic model simulates the numerical analysis performed multiple times iterations by Computational Fluid Dynamics (CFD). In addition, the flow analytic model can also include one or more signal generating models and one or more analytic models.

The flow analyzer **200** performs the flow analysis for the plurality of cells CE of the space around the flow analysis target component CP through the flow analytic model including one or more signal generating model and one or more analytic model derived from the model deriver **100** (operation S**120**).

The above-described operations S**110** and S**120** will be described in more detail.

**110**) according to an exemplary embodiment.

Referring to **120** constitutes a relationship equation where parameters, that is, a degree of influence H, a weight Q, and a cancellation constant D are not determined (operation S**210**). For example, the relationship equation where the parameters are not determined is as in the Equation 1 where the parameters H, Q, and D are unknown.

The signal generating model deriver **120** puts the analytic data into the relationship equation, and derives the parameters H, Q, and D of the relationship equation by using an optimization algorithm (operation S**220**). Herein, the optimization algorithm can be, for example, a Least-Squares Method, a Backpropagation algorithm, etc. For example, the analytic data can be V(k), Y(k), and V(k+T) used for the existing numerical analysis.

The signal generating model deriver **120** generates the signal generating model by applying the parameters H, Q, and D to the relationship equation (operation S**230**). For example, the signal generating model as in the Equation 1 is completed by applying the values of the parameters H, Q, and D to the relationship equation. This signal generating model predicts the input signal contributing to the output signal of the numerical analysis performed multiple times iterations.

The analytic model deriver **130** constitutes the relationship equation of the analytic model where a degree of influence A, a weight P, and a cancellation constant C that are parameters are not determined (operation S**240**). For example, the relationship equation where the parameters are not determined is as in the Equation 2 where the parameters A, P, and C are unknown.

The analytic model deriver **130** puts the analytic data into the relationship equation, and derives the parameters A, P, and C of the relationship equation through the optimization algorithm. For example, the analytic data can be Y(k), V(k+T), and Y(k+T) used for the existing numerical analysis (operation S**250**).

The analytic model deriver **130** generates the analytic model by applying the parameters A, P, and C to the relationship equation (operation S**260**). For example, the analytical model as in the Equation 2 is completed by applying the values of the parameters A, P, and C to the relationship equation. This analytical model predicts the output signal of the numerical analysis performed multiple times iterations.

As described above, the signal generating model is generated in the operation S**230**, and the analytic model is generated in the operation S**260**, thereby completing the flow analytic model including the signal generating model and the analytic model.

A method for performing the flow analysis by using the above-described flow analytic model will be described. **120**) according to an exemplary embodiment.

Referring to **210** derives an input signal and an output signal by the numerical analysis by performing the numerical analysis by Computational Fluid Dynamics (CFD) (operation S**310**). For example, according to the Equations 1 and 2, the input signal can be V(k) and the output signal can be Y(k).

The signal generator **220** predicts the input signal contributing to the output signal of the numerical analysis performed multiple times (k+T) iterations by reflecting the input signal and the output signal derived from the numerical analyzer **210** to the signal generating model (operation S**320**). For example, the signal generator **220** can derive the input signal {circumflex over (V)}(k+T) contributing to the output signal of the numerical analysis performed multiple times (k+T) iterations by inputting the input signal V(k) and the output signal Y(k) to the signal generating model as in the Equation 1.

The analyzer **230** predicts the output signal of the numerical analysis performed multiple times (k+T) iterations by reflecting the input signal predicted by the signal generator **220** and the output signal derived from the numerical analyzer **210** to the analytic model derived from the analytic model deriver **130** (operation S**330**). For example, the analyzer **230** can derive the output signal Ŷ(k+T) of the numerical analysis performed multiple times (k+T) iterations by inputting the predicted input signal {circumflex over (V)}(k+T) and output signal Y(k) to the analytical model as in the Equation 2.

**100** can be the apparatus described in the present specification (e.g., the flow analysis apparatus, etc.).

Referring to **100** can include at least one processor TN**110**, a transceiver TN**120**, and a memory TN**130**. In addition, the computing apparatus TN**100** can further include a storage device TN**140**, an input interface TN**150**, an output interface TN**160**, etc. The components included in the computing apparatus TN**100** can be connected by a bus TN**170** and communicate with each other.

The processor TN**110** can execute a program command stored in at least one of the memory TN**130** and the storage device TN**140**. The processor TN**110** can include a central processing unit (CPU), a graphics processing unit (GPU), or a dedicated processor on which the methods according to an exemplary embodiment are performed. The processor TN**110** can be configured to implement the procedures, functions, methods, etc. described in connection with an exemplary embodiment. The processor TN**110** can control each component of the computing apparatus TN**100**.

Each of the memory TN**130** and the storage device TN**140** can store various information related to an operation of the processor TN**110**. Each of the memory TN**130** and the storage device TN**140** can be composed of at least one of a volatile storage medium and a nonvolatile storage medium. For example, the memory TN**130** can be composed of at least one of a read only memory (ROM) and a random access memory (RAM).

The transceiver TN**120** can transmit and/or receive a wired signal or a wireless signal. The transceiver TN**120** can be connected to a network to perform communication.

Meanwhile, the flow analysis method according to an exemplary embodiment can be implemented in the form of a readable program through various computer means and recorded in a computer-readable recording medium. Herein, the recording medium can include program commands, data files, data structures, etc. alone or in combination thereof. The program commands to be recorded on the recording medium can be those specially designed and constructed for the present disclosure or can also be those known and available to those skilled in the art of computer software. For example, the recording medium can be magnetic media such as hard disks, floppy disks, and magnetic tapes, optical media such as CD-ROMs and DVDs, magneto-optical media such as floptical disks, and hardware devices specially configured to store and execute the program commands such as ROMs, RAMs, and flash memory. Examples of the program commands can include not only machine language wires such as those produced by a compiler but also high-level language wires that can be executed by a computer by using an interpreter, etc. This hardware device can be configured to operate as one or more software modules in order to perform the operation of the present disclosure, and vice versa.

While one or more exemplary embodiments have been described with reference to the accompanying drawings, it is to be understood by those skilled in the art that various modifications and changes in form and details can be made therein without departing from the spirit and scope as defined by the appended claims. Therefore, the description of the exemplary embodiments should be construed in a descriptive sense only and not to limit the scope of the claims, and many alternatives, modifications, and variations will be apparent to those skilled in the art.

## Claims

1. A flow analysis apparatus, comprising:

- a model deriver configured to generate a flow analytic model for performing a flow analysis for a plurality of cells by simulating a numerical analysis by Computational Fluid Dynamics (CFD) with respect to the plurality of cells that divide a space around a component; and

- a flow analyzer configured to perform the flow analysis for the plurality of cells that divide the space around a design target component by using the generated flow analytic model.

2. The flow analysis apparatus of claim 1,

- wherein the model deriver comprises:

- an analyzing data storage configured to store analytic data used for the numerical analysis;

- a signal generating model deriver configured to generate a signal generating model for predicting an input signal contributing to an output signal of the numerical analysis performed multiple times iterations through the analytic data; and

- an analytic model deriver configured to generate the analytic data for predicting the output signal of the numerical analysis performed multiple times iterations through the analytic data.

3. The flow analysis apparatus of claim 2, V ^ ( k + T ) = HQ [ V ( k ) Y ( k ) ] + D,

- wherein the signal generating model deriver predicts the input signal contributing to the output signal of the numerical analysis performed multiple times iterations through an Equation

- wherein the k and the T refer to the number of times of numerical analysis,

- wherein the H refers to a degree of influence,

- wherein the Q refers to a weight,

- wherein the D refers to a cancellation constant,

- wherein the V(k) refers to an input signal of the kth numerical analysis,

- wherein the Y(k) refers to an output signal of the kth numerical analysis, and

- wherein the {circumflex over (V)}(k+T) refers to an input signal of the (k+T)th numerical analysis.

4. The flow analysis apparatus of claim 2,

- wherein the signal generating model deriver generates a signal generating model by deriving a parameter through an optimization algorithm after constituting a relationship equation of the signal generating model where the parameter is not determined.

5. The flow analysis apparatus of claim 2, Y ^ ( k + T ) = AP [ Y ( k ) V ^ ( k + T ) ] + C,

- wherein the analytic model deriver predicts the output signal of the numerical analysis performed multiple times iterations through an Equation

- wherein the k and the T refer to the number of times of numerical analysis,

- wherein the A refers to a degree of influence,

- wherein the P refers to a weight,

- wherein the C refers to a cancellation constant,

- wherein the Y(k) refers to an output signal of the kth numerical analysis,

- wherein the {circumflex over (V)}(k+T) refers to an input signal of the (k+T)th numerical analysis, and

- wherein the Ŷ(k+T) refers to an output signal of the (k+T)th numerical analysis.

6. The flow analysis apparatus of claim 2,

- wherein the analytic model deriver generates an analytic model by deriving a parameter through an optimization algorithm after constituting a relationship equation of the analytic model where the parameter is not determined.

7. The flow analysis apparatus of claim 2,

- wherein the flow analyzer comprises:

- a numerical analyzer configured to derive the analytic data by performing the numerical analysis for the design target component;

- a signal generator configured to derive the input signal contributing to the output signal of the numerical analysis performed multiple times iterations through the signal generating model derived from the signal generating model deriver and the analytic data derived from the numerical analyzer; and

- an analyzer configured to derive the output signal of the numerical analysis performed multiple times iterations through the analytic model derived from the analytic model deriver from the input signal predicted by the signal generator.

8. The flow analysis apparatus of claim 7,

- wherein the numerical analyzer derives the analytic data by performing the numerical analysis by Computational Fluid Dynamics with respect to the plurality of cells that divide the space around the design target component.

9. The flow analysis apparatus of claim 8,

- wherein the signal generator derives the input signal contributing to the output signal of the numerical analysis performed a predetermined number of times iterations by inputting the analytic data to the signal generating model.

10. The flow analysis apparatus of claim 9,

- wherein the analyzer derives the output signal where the numerical analysis has been performed the predetermined number of times iterations by inputting the output signal of the analytic data and the contributing input signal to the analytic model.

11. A flow analysis apparatus, comprising:

- a model deriver configured to generate a flow analytic model for performing a flow analysis for a plurality of cells by using analytic data used for a numerical analysis by Computational Fluid Dynamics (CFD) with respect to the plurality of cells that divide a space around a component; and

- a flow analyzer configured to perform the flow analysis for the plurality of cells that divide the space around a design target component by using the generated flow analytic model.

12. The flow analysis apparatus of claim 11,

- wherein the flow analytic model simulates the numerical analysis performed multiple times iterations by the Computational Fluid Dynamics.

13. The flow analysis apparatus of claim 12,

- wherein the flow analytic model comprises:

- one or more signal generating models for predicting an input signal contributing to a result of the numerical analysis performed multiple times iterations among the plurality of input signals through the analytic data; and

- an analytic model for predicting the result of the numerical analysis performed multiple times iterations through the analytic data.

14. The flow analysis apparatus of claim 13, V ^ ( k + T ) = HQ [ V ( k ) Y ( k ) ] + D,

- wherein the signal generating model predicts the input signal contributing to the output signal of the numerical analysis performed multiple times iterations through an Equation

- wherein the k and the T refer to parameters representing the number of times of numerical analysis,

- wherein the H refers to a degree of influence,

- wherein the Q refers to a weight,

- wherein the D refers to a cancellation constant,

- wherein the V(k) refers to an input signal of the kth numerical analysis,

- wherein the Y(k) refers to an output signal of the kth numerical analysis, and

- wherein the {circumflex over (V)}(k+T) refers to an input signal of the (k+T)th numerical analysis.

15. The flow analysis apparatus of claim 13, Y ^ ( k + T ) = AP [ Y ( k ) V ^ ( k + T ) ] + C,

- wherein the analytic model predicts the output signal of the numerical analysis performed multiple times iterations through an Equation

- wherein the k and the T refer to the number of times of numerical analysis,

- wherein the A refers to a degree of influence,

- wherein the P refers to a weight,

- wherein the C refers to a cancellation constant,

- wherein the Y(k) refers to an output signal of the kth numerical analysis,

- wherein the {circumflex over (V)}(k+T) refers to an input signal of the (k+T)th numerical analysis, and

- wherein the Ŷ(k) refers to an output signal of the (k+T)th numerical analysis.

16. A flow analysis method, comprising:

- storing, by a model deriver, analytic data comprising a plurality of input signals used for a numerical analysis by Computational Fluid Dynamics (CFD) with respect to a plurality of cells that divide a space around a component and a plurality of output signals corresponding to each of the plurality of input signals;

- generating, by the model deriver, a flow analytic model for performing a flow analysis for the plurality of cells by using the analytic data; and

- performing, by a flow analyzer, the flow analysis for the plurality of cells that divide the space around a design target component by using the flow analytic model.

17. The flow analysis method of claim 16,

- wherein the generating the flow analytic model comprises:

- generating, by a signal generating model deriver, a signal generating model for predicting an input signal contributing to an output signal of the numerical analysis performed multiple times iterations among the plurality of input signals through the analytic data; and

- generating, by an analytic model deriver. the analytic model for predicting the output signal of the numerical analysis performed multiple times iterations among the plurality of input signals through the analytic data.

18. The flow analysis method of claim 17, V ^ ( k + T ) = HQ [ V ( k ) Y ( k ) ] + D,

- wherein the generating the signal generating model comprises predicting the input signal contributing to the output signal of the numerical analysis performed multiple times iterations through an Equation

- wherein the k and the T refer to the number of times of numerical analysis,

- wherein the H refers to a degree of influence,

- wherein the Q refers to a weight,

- wherein the D refers to a cancellation constant,

- wherein the V(k) refers to an input signal of the kth numerical analysis,

- wherein the Y(k) refers to an output signal of the kth numerical analysis, and

- wherein the {circumflex over (V)}(k+T) refers to an input signal of the (k+T)th numerical analysis.

19. The flow analysis method of claim 17, Y ^ ( k + T ) = AP [ Y ( k ) V ^ ( k + T ) ] + C,

- wherein the generating the analytic model comprises predicting the output signal of the numerical analysis performed multiple times iterations through an Equation

- wherein the k and the T refer to the number of times of numerical analysis,

- wherein the A refers to a degree of influence,

- wherein the P refers to a weight,

- wherein the C refers to a cancellation constant,

- wherein the Y(k) refers to an output signal of the kth numerical analysis,

- wherein the {circumflex over (V)}(k+T) refers to an input signal of the (k+T)th numerical analysis, and

- wherein the Ŷ(k+T) refers to an output signal of the (k+T)th numerical analysis.

20. The flow analysis method of claim 17,

- wherein the generating the signal generating model comprises:

- constituting, by the signal generating model deriver, a relationship equation of the signal generating model where a parameter is not determined; and

- completing, by the signal generating model deriver, the signal generating model by deriving the parameter through an optimization algorithm.

**Patent History**

**Publication number**: 20200065448

**Type:**Application

**Filed**: May 23, 2019

**Publication Date**: Feb 27, 2020

**Inventors**: Jeehun PARK (Gwangmyeong-si), Jaehyeon PARK (Hwaseong-si), Sangjin LEE (Yongin-si), Hyunsik KIM (Gimpo-si)

**Application Number**: 16/420,179

**Classifications**

**International Classification**: G06F 17/50 (20060101);