KR20190061616A - Method and apparatus for simulation model optimization using component-based dynamic model reconstruction - Google Patents
Method and apparatus for simulation model optimization using component-based dynamic model reconstruction Download PDFInfo
- Publication number
- KR20190061616A KR20190061616A KR1020170160180A KR20170160180A KR20190061616A KR 20190061616 A KR20190061616 A KR 20190061616A KR 1020170160180 A KR1020170160180 A KR 1020170160180A KR 20170160180 A KR20170160180 A KR 20170160180A KR 20190061616 A KR20190061616 A KR 20190061616A
- Authority
- KR
- South Korea
- Prior art keywords
- simulation model
- optimization
- component
- combination
- model
- Prior art date
Links
Images
Classifications
-
- G06F17/5009—
Landscapes
- Management, Administration, Business Operations System, And Electronic Commerce (AREA)
Abstract
Description
The present invention relates to simulations, and more particularly, to a method and apparatus for optimizing a simulation model utilizing component-based dynamic model reconstruction.
Modeling and simulation are widely used to understand, analyze, and design complex systems in many areas. This is due to the development of modeling and simulation methods. In other words, it is changing from past ad hoc modeling and simulation methods to systematic development based on system theory. The development of these modeling and simulation technologies is based on the idea that a model is developed at a low cost by clearly distinguishing between a simulation model and an engine and a complex model is expressed as a combination of modular components, Increase understanding and analysis, and increase opportunities to reuse existing models during development. This reduces the time and cost of the overall modeling and simulation technology and provides a theoretical and practical tool for efficiently analyzing complex systems that can not be analyzed by hermetic methods. An example of such a technology is the Discrete Event System Specification (DEVS) formalism, which is currently being used to develop simulation models in various fields both at home and abroad.
On the other hand, the core of modeling and simulation technology based on system theory is a method of modular component development that can express the dynamics of the whole system through the separation of model and simulation engine and the combination of components constituting the system. The separation of the model and the simulation engine and the combination of the components that make up the system enable the development of the simulation framework and provide an environment in which various components can be freely configured within the framework. Model Dynamic reconfiguration refers to changing the combination of components that make up the model during simulation execution. In other words, the number of components or the connection method of the components can be changed during the simulation, which is based on the function of managing or modifying information of a definite model component. Therefore, model dynamic reconfiguration can be supported in an environment with a theoretical basis of modeling and simulation, such as formalism.
Model optimization represents the task of finding the optimal combination of parameters of a model in the direction of increasing the performance of the model. In this process, various methods are used to analyze the performance of the model. Simulation-based optimization analyzes the performance of a simulation model as a simulation result and shows how to find a combination of parameters of the model for good performance. Especially, it is difficult to apply the general analytical method to the analysis of a model having complex and nonlinear characteristics. However, it is advantageous to solve this problem by simulation analysis. In recent years, along with the development of machine learning and artificial intelligence technologies, simulation-based optimization has also been developed in various ways.
However, the existing simulation-based optimization is centered on the model parameters, and there is a disadvantage that the optimization of the model beyond the range that the parameter can not express can not be achieved.
A problem to be solved by the present invention is to provide a method and apparatus for achieving model optimization using model parameters and dynamic model reconstruction.
A method according to an embodiment of the present invention is a method for a simulation model optimization, the method comprising: setting an optimization target; The apparatus comprising: initially executing a predetermined simulation model; The apparatus comprising: comparing the result of the simulation model execution with the optimization target; Searching for a combination of components suitable for the simulation model if the result of the execution of the simulation model does not satisfy the optimization target and searching for a combination of parameters suitable for the simulation model; Reconstructing the simulation model based on the searched component combination and parameter combination; And executing the reconstructed simulation model, wherein an optimal simulation model is obtained based on a comparison between the result of executing the reconstructed simulation model and the optimization target.
Determining a combination of the parameters until a result of the execution of the simulation model satisfies the optimization target, reconstructing the simulation model, executing the reconstructed simulation model, and comparing the optimization goals Can be repeatedly performed.
The reconstructing of the simulation model may perform a dynamic exchange of components of the simulation model based on the searched component combination to reconstruct the simulation model.
The step of finding the combination of parameters may comprise the step of identifying, from a database comprising a module having a module having an independent input / output interface, a component to be utilized by clustering the component on the basis of input and output interfaces have.
The finding of the combination of parameters may comprise generating a new component for achieving the optimization goal using the optimization goal, existing optimization result data, and the components that make up the simulation model.
Wherein reconfiguring the simulation model further comprises reconfiguring the simulation model using the newly generated component, wherein the optimization result data includes optimization history including component and parameter information for a simulation model that achieves a random optimization goal Information.
According to another aspect of the present invention, there is provided an apparatus for optimizing a simulation model, comprising: an input / output unit configured to receive data; And a processor coupled to the input / output unit and configured to perform a simulation model optimization, wherein the processor is configured to perform the optimization of the simulation model when the result of executing the simulation model that is initially set in advance does not satisfy the preset optimization target Searching for a combination of components suitable for the simulation model, and reconstructing and executing the simulation model based on the searched component combination and parameter combination, and comparing the result of executing the reconstructed simulation model and the optimization An optimal simulation model is obtained based on the comparison of the targets.
Wherein the processor is further configured to find a combination of the component and parameter suitable for the simulation model and reconstruct and execute the simulation model until the result of the execution of the simulation model satisfies the optimization target, And comparing the targets to obtain the optimal simulation model.
The processor may be further configured to perform a dynamic exchange of components of the simulation model based on the searched component combination to reconstruct the simulation model.
The apparatus may further include a component database storing components composed of modules having independent input / output interfaces. The processor may be configured to cluster components based on input and output interfaces to identify components suitable for the simulation model from the component database.
The apparatus may further include an optimization history database storing optimization result data, which is optimization history information including component and parameter information for a simulation model that achieves a certain optimization target. At this time, the processor generates a new component for achieving the optimization target by using the optimization target, the optimization result data, and the components that configure the simulation model, Lt; / RTI >
The processor comprising: a model optimization estimator configured to execute the simulation model and to perform a model optimization evaluation by comparing the execution result and the optimization target; An optimal combination navigator configured to obtain optimal combination of components and parameters to achieve the optimization goal; A dynamic model combiner for reconstructing the simulation model based on the combination of components and parameters; And an optimal component generator configured to newly generate a component required to achieve the optimization target.
According to the embodiment of the present invention, by applying the model parameters and the dynamic model reconstruction method, a wider range of model optimization is performed and a new component is presented using the existing optimization result, Can be achieved.
In particular, according to embodiments of the present invention, simulation-based model optimization can be improved in two directions.
First, a variety of simulation-based model optimization methods are provided. In the conventional method, we tried to optimize the model by searching for a combination of various parameters of the simulation model. However, a case where the parameter combination does not achieve a certain level or more of optimization is uttered. Therefore, in the embodiment of the present invention, the search range of the parameter and the component model is searched to consider the higher dimensional search range, and in particular, the simulation execution time is shortened through the dynamic component exchange.
Second, when the combination of the parameter and the existing component model is insufficient, a function of presenting a new component in consideration of the existing optimization history is provided. Thus, it is expected that optimization of various simulation models can be achieved more efficiently.
1 is a structural diagram of a simulation model optimizing apparatus according to an embodiment of the present invention.
2 is a flowchart of a simulation model optimization method according to an embodiment of the present invention.
3 is an operation diagram illustrating that an optimal model acquisition process and an optimal component process are performed in the simulation model optimizing apparatus according to the embodiment of the present invention.
FIG. 4 is a diagram illustrating a process of searching and confirming a component in a DB that can be utilized in model reconstruction in an optimal model combiner according to an embodiment of the present invention.
5 is a diagram illustrating component registration and addition in accordance with an embodiment of the present invention.
6 is a structural diagram of a simulation model optimizing apparatus according to another embodiment of the present invention.
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. In order to clearly illustrate the present invention, parts not related to the description are omitted, and similar parts are denoted by like reference characters throughout the specification.
Throughout the specification, when an element is referred to as " comprising ", it means that it can include other elements as well, without excluding other elements unless specifically stated otherwise.
Hereinafter, a method and apparatus for optimizing a simulation model using a component-based dynamic model reconfiguration according to an embodiment of the present invention will be described.
1 is a diagram illustrating a structure of a simulation model optimizing apparatus according to an embodiment of the present invention.
1, a simulation
The
The
The dynamic model combiner 130 constitutes a simulation model. The
The
The
The
In the embodiment of the present invention, based on the simulation model optimizing apparatus having such a structure, an optimal model acquiring process for acquiring an optimum model for achieving a given optimization goal, and a model acquiring process for acquiring model information, optimization history, , And performs an optimal component generation process to generate a component capable of achieving the current optimization goal.
2 is a flowchart of a simulation model optimization method according to an embodiment of the present invention. 3 is an operation diagram illustrating that an optimal model acquisition process and an optimal component process are performed in the simulation model optimizing apparatus according to the embodiment of the present invention.
As shown in FIGS. 2 and 3, the simulation
To this end, an optimization target is first set (S100). Through the user interface, the optimization target required by the user is input to the
The
When the initial simulation model is executed and a result is generated, the
FIG. 4 is a diagram illustrating a process of searching and confirming a component in a DB that can be utilized in model reconstruction in an optimal model combiner according to an embodiment of the present invention.
The components stored in the
On the other hand, the model has several parameters, and each parameter has a range of values. The optimum parameter combination is obtained by analyzing the relation between the parameter combination and the simulation result based on the result of executing the simulation by variously changing the combination of each parameter value to find the optimum parameter combination that matches the optimization target of the model. Design of experiments can be applied to this process, and factorial design and response surface design can be applied.
As described above, when the optimal component combination and parameter combination is searched, the
Based on the new simulation model configured by the
In accordance with the repetition of this process, when the simulation result E reaches the optimization target?, The optimum model is obtained (S160). Information on the simulation model can be stored and managed in the
According to the embodiment of the present invention, the optimization of the simulation model can be performed by searching the combination of the parameter and the component model and considering the higher-dimensional search range. This provides a variety of model optimization methods.
Meanwhile, in the embodiment of the present invention, various simulation model optimization can be performed by presenting a new component in consideration of an optimization history.
5 is a diagram illustrating component registration and addition in accordance with an embodiment of the present invention.
There are two ways to register a component in the component DB. Basically, the first method is a method in which a user directly creates a component and registers it as an API (application program interface) provided by the component DB. The second method is a method of registering a new component through generation of an optimal component according to an embodiment of the present invention.
In an embodiment of the present invention, using the initial optimization goals, existing optimization result data (optimization result data stored in the optimization history DB 151), and the current component, the
Thus, the newly generated component by the
6 is a structural diagram of a simulation model optimizing apparatus according to another embodiment of the present invention.
6, the simulation
The
The
The input /
The embodiments of the present invention are not limited to the above-described apparatuses and / or methods, but may be implemented through a program for realizing functions corresponding to the configuration of the embodiment of the present invention, a recording medium on which the program is recorded And such an embodiment can be easily implemented by those skilled in the art from the description of the embodiments described above.
While the present invention has been particularly shown and described with reference to exemplary embodiments thereof, it is to be understood that the invention is not limited to the disclosed exemplary embodiments, It belongs to the scope of right.
Claims (12)
The apparatus comprising: setting an optimization target;
The apparatus comprising: initially executing a predetermined simulation model;
The apparatus comprising: comparing the result of the simulation model execution with the optimization target;
Searching for a combination of components suitable for the simulation model if the result of the execution of the simulation model does not satisfy the optimization target and searching for a combination of parameters suitable for the simulation model;
Reconstructing the simulation model based on the searched component combination and parameter combination; And
Executing the reconstructed simulation model
Lt; / RTI >
And an optimal simulation model is obtained based on a comparison of the result of executing the reconstructed simulation model and the optimization goal.
Determining a combination of the parameters until a result of the execution of the simulation model satisfies the optimization target, reconstructing the simulation model, executing the reconstructed simulation model, and comparing the optimization goals Lt; / RTI >
Wherein reconstructing the simulation model comprises:
And performing a dynamic exchange of components of the simulation model based on the searched component combination to reconstruct the simulation model.
Wherein the step of finding a combination of the parameters comprises:
Identifying a component to be utilized by clustering components based on input and output interfaces from a database of components comprising modules having independent input and output interfaces;
/ RTI >
Wherein the step of finding a combination of the parameters comprises:
Generating a new component for achieving the optimization goal using the optimization goal, existing optimization result data, and the components that make up the simulation model
/ RTI >
Wherein reconstructing the simulation model comprises reconstructing the simulation model using the newly generated component,
Wherein the optimization result data is optimization history information including component and parameter information for a simulation model that achieves any optimization goal.
An input / output unit configured to receive input of data; And
A processor coupled to the input / output unit and configured to perform simulation model optimization,
The processor comprising:
Searching for a combination of components suitable for the simulation model and searching for a combination of parameters suitable for the simulation model when the result of executing the initial simulation model does not satisfy a preset optimization target, And reconstruct and run the simulation model based on the combination,
And an optimal simulation model is obtained based on a comparison of the result of executing the reconstructed simulation model and the optimization target.
Wherein the processor is further configured to find a combination of the component and parameter suitable for the simulation model and reconstruct and execute the simulation model until the result of the execution of the simulation model satisfies the optimization target, And to perform the process of comparing the targets repeatedly to obtain the optimal simulation model.
Wherein the processor is further configured to perform a dynamic exchange of components of the simulation model based on the searched component combination to reconstruct the simulation model.
A component database consisting of modules having independent input / output interfaces is stored
Further comprising:
Wherein the processor is configured to cluster components based on input and output interfaces to identify components suitable for the simulation model from the component database.
An optimization history database storing optimization result data, which is optimization history information including component and parameter information for a simulation model that achieves a random optimization target
Further comprising:
The processor comprising:
And generating a new component for achieving the optimization goal using the optimization target, the optimization result data, and a component that constitutes the simulation model, and reconstructing the simulation model using the newly generated component, Device.
The processor comprising:
A model optimization estimator configured to execute the simulation model and to perform a model optimization evaluation by comparing the execution result and the optimization target;
An optimal combination navigator configured to obtain optimal combination of components and parameters to achieve the optimization goal;
A dynamic model combiner for reconstructing the simulation model based on the combination of components and parameters; And
An optimal component generator configured to newly generate a component required to achieve the optimization target,
. ≪ / RTI >
Priority Applications (1)
Application Number | Priority Date | Filing Date | Title |
---|---|---|---|
KR1020170160180A KR20190061616A (en) | 2017-11-28 | 2017-11-28 | Method and apparatus for simulation model optimization using component-based dynamic model reconstruction |
Applications Claiming Priority (1)
Application Number | Priority Date | Filing Date | Title |
---|---|---|---|
KR1020170160180A KR20190061616A (en) | 2017-11-28 | 2017-11-28 | Method and apparatus for simulation model optimization using component-based dynamic model reconstruction |
Publications (1)
Publication Number | Publication Date |
---|---|
KR20190061616A true KR20190061616A (en) | 2019-06-05 |
Family
ID=66845117
Family Applications (1)
Application Number | Title | Priority Date | Filing Date |
---|---|---|---|
KR1020170160180A KR20190061616A (en) | 2017-11-28 | 2017-11-28 | Method and apparatus for simulation model optimization using component-based dynamic model reconstruction |
Country Status (1)
Country | Link |
---|---|
KR (1) | KR20190061616A (en) |
Cited By (4)
Publication number | Priority date | Publication date | Assignee | Title |
---|---|---|---|---|
CN112612772A (en) * | 2020-12-01 | 2021-04-06 | 中车长江车辆有限公司 | Polyurethane foaming simulation database construction method and device |
KR20210117030A (en) * | 2020-03-18 | 2021-09-28 | 인하대학교 산학협력단 | Digital twin system and method for virtualization of autonomous driving |
CN114861403A (en) * | 2022-04-12 | 2022-08-05 | 国网江苏省电力有限公司电力科学研究院 | Dynamic simulation step length optimization method and system for electrothermal coupling network |
CN117473932A (en) * | 2023-10-27 | 2024-01-30 | 华南理工大学 | Agile design library driven DTCO efficient optimization method |
-
2017
- 2017-11-28 KR KR1020170160180A patent/KR20190061616A/en unknown
Cited By (6)
Publication number | Priority date | Publication date | Assignee | Title |
---|---|---|---|---|
KR20210117030A (en) * | 2020-03-18 | 2021-09-28 | 인하대학교 산학협력단 | Digital twin system and method for virtualization of autonomous driving |
CN112612772A (en) * | 2020-12-01 | 2021-04-06 | 中车长江车辆有限公司 | Polyurethane foaming simulation database construction method and device |
CN112612772B (en) * | 2020-12-01 | 2023-07-18 | 中车长江车辆有限公司 | Database construction method and device for polyurethane foaming simulation |
CN114861403A (en) * | 2022-04-12 | 2022-08-05 | 国网江苏省电力有限公司电力科学研究院 | Dynamic simulation step length optimization method and system for electrothermal coupling network |
CN114861403B (en) * | 2022-04-12 | 2023-10-27 | 国网江苏省电力有限公司电力科学研究院 | Electric heating coupling network dynamic simulation step length optimization method and system |
CN117473932A (en) * | 2023-10-27 | 2024-01-30 | 华南理工大学 | Agile design library driven DTCO efficient optimization method |
Similar Documents
Publication | Publication Date | Title |
---|---|---|
AU2019200046B2 (en) | Utilizing artificial intelligence to test cloud applications | |
CN109542791B (en) | A kind of program large-scale concurrent evaluating method based on container technique | |
Appice et al. | A co-training strategy for multiple view clustering in process mining | |
KR20190061616A (en) | Method and apparatus for simulation model optimization using component-based dynamic model reconstruction | |
Calinescu et al. | Formal methods@ runtime | |
US20210165640A1 (en) | Accelerating Application Modernization | |
Prats et al. | You only run once: spark auto-tuning from a single run | |
CN109783459A (en) | The method, apparatus and computer readable storage medium of data are extracted from log | |
Lin et al. | Adaptive code learning for spark configuration tuning | |
US20090164501A1 (en) | E-matching for smt solvers | |
CN112949711B (en) | Neural network model multiplexing training method and device for software defined satellites | |
US20180309854A1 (en) | Protocol model generator and modeling method thereof | |
Swain et al. | Test case design using slicing of UML interaction diagram | |
US20230333971A1 (en) | Workload generation for optimal stress testing of big data management systems | |
Ostrowski et al. | Knowledge-based software testing agent using evolutionary learning with cultural algorithms | |
Berral et al. | ALOJA: a framework for benchmarking and predictive analytics in Hadoop deployments | |
Pellegrini et al. | Optimizing mpi runtime parameter settings by using machine learning | |
CN114564523B (en) | Big data vulnerability analysis method and cloud AI system for intelligent virtual scene | |
EP4160435A1 (en) | Cleaning raw data generated by a telecommunications network for deployment in a deep neural network model | |
Hsu et al. | Low-level augmented bayesian optimization for finding the best cloud vm | |
Islam et al. | A Package Based Clustering for enhancing software defect prediction accuracy | |
Goyal et al. | Smartphone context event sequence prediction with poermh and tke-rules algorithms | |
Shchekotykhin et al. | On computing minimal conflicts for ontology debugging | |
Du | Opaque response generation enabling automatic creation of virtual services for service virtualisation | |
CN116527411B (en) | Data security intelligent protection model construction method and device and collaboration platform |