연구에서 모형의 분류와 검정모형의 개발

권순달(Soon-Dal Kwon) 한국교육학회 2023 敎育學硏究 Vol.61 No.1

본 연구는 연구에서 널리 활용되고 있는 모형을 분류하고, 다양한 모형을 이해하는 데 있다. 이를 위하여 모형의 개념을 재정의하고, 모형의 분류 준거를 분석하고, 새로운 모형 분류 기준을 제시하였다. 본 연구에서 모형이란, “연구자의 이론, 철학, 신념 등을 반영한 것으로 논리적으로 제안된 그 무엇이다”라고 정의하였다. 여러 학문 분야의 모형 분류 준거가 다양하고, 대부분 이분화된 준거를 사용하였다. 이에 따라 연구자도 검정가능성을 모형의 분류 준거로 하여 검정모형과 비검정모형으로 분류하였다. 이 준거에 의하면 검정모형 제안자들은 데이터를 사용하여 자신이 개발한 모형을 검정하여 제시하고 있다. 그러나 비검정모형은 엄격한 검정이 이루어지 않고 제안되었다고 할 수 있다. 그리고 비검정모형들은 개발과정의 일반적인 절차를 따르지 않고 있다. 이러한 이유에서 검정모형 개발 5단계를 제안하였다. 물론 연구자가 제안한 검정 모형 개발 5단계는 하나의 ‘’비검정 모형‘이라고 할 수 있다. 검정모형과 비검정 모형의 개발은 누구나 할 수 있지만, 비검정모형이 아닌, 검정모형은 논리적인 절차에 따라 개발되고 검정될 필요가 있다. The purpose for this study is to classify models widely used in research and to understand various models. To this end, the concept of the model was redefined, the classification criteria of models were analyzed, and new model classification criteria were presented. In this study, a model is defined as “something that is logically proposed as a reflection of the researcher's theory, philosophy, beliefs, etc.”. There are various criteria for classification of models in many academic fields, but most of them use the dichotomous method. Accordingly, this researcher classified all models into a tested model and the non-tested model based on the test feasibility. The tested model refers to a model tested by the proposers using data. On the other side, nnn-tested model is a model proposed without rigorous testing. Tested models require rigorous development procedures, but non-tested models do not require such procedures. For this reason, five stages of test model development were proposed. Of course, the 5 stages of test model development proposed by the researcher can be said to be a ‘non-testing model’. Anyone can develop a model whether it is a test model or a non-test model, but a test model, not a non-test model, needs to be developed and tested strictly.

모델 기반 평가 방법을 이용한 지능형자동차 기능의 평가 자동화에 관한 연구

한갑수(Kab-Su Han),강호준(Ho-joon Kang),정승환(Seung-Hwan Chung),권성진(Seong-Jin Kwon),이봉현(Bong-Hyun Lee) 한국자동차공학회 2013 한국자동차공학회 부문종합 학술대회 Vol.2013 No.5

Model-based testing is a kind of black box testing that test suites are derived from behavioral model of SUT through test case generation and execute test suites automatically by model-based testing tool. A model-based testing tool uses various test generation algorithms and criteria to generate test cases. The behavioral model of SUT has to be concise enough to write and validate. Also, be precise enough to specify the behavior which is to be tested. Test cases can be generated automatically from the model and test engineer can control testing and maintain test cases easily. The abstract test cases have to be transformed into executable test cases with more specific information about SUT. Test selection criteria and lots of test generation strategies are commonly used in model-based testing. The complementary of these are key point of model-based testing. For this purpose, several state-based models are developed and test suites which derived from various test generation strategies, especially focused on transition coverage, are performed. Also, testing performances are discussed for test coverage of each test suite.

A Comparison Among Neo-Hookean Model, Mooney-Rivlin Model, and Ogden Model for Chloroprene Rubber

김범근,이성범,이재원,조세현,박형민,염상훈,박성한 한국정밀공학회 2012 International Journal of Precision Engineering and Vol. No.

Neo-Hookean model and Mooney-Rivlin model are hyperelastic material models where the strain energy density function is made from invariants of the left Cauchy-Green deformation tensor. Even though Ogden model is a hyperelastic material model, its strain energy density function is expressed by principal stretch ratio. These three models have been widely used in industries. Recently, Ogden model, especially Ogden 3rd model, shows better agreement with the test data than others. In spite of the limitations to describe particular stress states, it is known that reasonable results using these models can be obtained for various structural components. In this research, three kinds of models are considered for Chloroprene rubber. Three kinds of tests (Uniaxial tension test, Biaxial tension test, and Planar shear test) are performed for Chloroprene specimen and through four kinds of test combinations (Uni+Bi, Uni+Pl, Bi+Pl, Uni+Bi+Pl), numerical simulations are carried out for Neo-Hookean model, Mooney-Rivlin model, and Ogden model. It is shown that Mooney-Rivlin model and Ogden model can be used for Chloroprene Rubber in the specific ranges for Isotropic Hyperelastic model.

Validation of a 750 kW semi-submersible floating offshore wind turbine numerical model with model test data, part II: Model-II

Kim, Junbae,Shin, Hyunkyoung The Society of Naval Architects of Korea 2020 International Journal of Naval Architecture and Oc Vol.12 No.1

Floating Offshore Wind Turbines (FOWT) installed in the deep sea regions where stable and strong wind flows are abundant would have significantly improved energy production capacity. When designing FOWT, it is essential to understand the stability and motion performance of the floater. Water tank model tests are required to evaluate these aspects of performance. This paper describes a model test and numerical simulation for a 750-kW semi-submersible platform wind turbine model-II. In the previous model test, the 750-kW FOWT model-I suffered slamming phenomena from extreme wave conditions. Because of that, the platform freeboard of model-II was increased to mitigate the slamming load on the platform deck structure in extreme conditions. Also, the model-I pitch Response Amplitude Operators (RAO) of simulation had strong responses to the natural frequency region. Thus, the hub height of model-II was decreased to reduce the pitch resonance responses from the low-frequency response of the system. Like the model-I, 750-kW FOWT model-II was built with a 1/40 scale ratio. Furthermore, the experiments to evaluate the performance characteristics of the model-II wind turbine were executed at the same location and in the same environment conditions as were those of model-I. These tests included a free decay test, and tests of regular and irregular wave conditions. Both the experimental and simulation conditions considered the blade rotating effect due to the wind. The results of the model tests were compared with the numerical simulations of the FOWT using FAST (Fatigue, Aerodynamics, Structures, and Turbulence) code from the National Renewable Energy Laboratory (NREL).

Validation of a 750 kW semi-submersible floating offshore wind turbine numerical model with model test data, part II: Model-II

Kim, Junbae,Shin, Hyunkyoung The Society of Naval Architects of Korea 2020 International Journal of Naval Architecture and Oc Vol.12 No.-

Floating Offshore Wind Turbines (FOWT) installed in the deep sea regions where stable and strong wind flows are abundant would have significantly improved energy production capacity. When designing FOWT, it is essential to understand the stability and motion performance of the floater. Water tank model tests are required to evaluate these aspects of performance. This paper describes a model test and numerical simulation for a 750-kW semi-submersible platform wind turbine model-II. In the previous model test, the 750-kW FOWT model-I suffered slamming phenomena from extreme wave conditions. Because of that, the platform freeboard of model-II was increased to mitigate the slamming load on the platform deck structure in extreme conditions. Also, the model-I pitch Response Amplitude Operators (RAO) of simulation had strong responses to the natural frequency region. Thus, the hub height of model-II was decreased to reduce the pitch resonance responses from the low-frequency response of the system. Like the model-I, 750-kW FOWT model-II was built with a 1/40 scale ratio. Furthermore, the experiments to evaluate the performance characteristics of the model-II wind turbine were executed at the same location and in the same environment conditions as were those of model-I. These tests included a free decay test, and tests of regular and irregular wave conditions. Both the experimental and simulation conditions considered the blade rotating effect due to the wind. The results of the model tests were compared with the numerical simulations of the FOWT using FAST (Fatigue, Aerodynamics, Structures, and Turbulence) code from the National Renewable Energy Laboratory (NREL).

이산 사건 시스템 명세 시뮬레이션에서 원자 모델의 빠른 검증을 위한 스크립트 기반 테스트 시스템

남수만(Su-Man Nam) 한국컴퓨터정보학회 2022 韓國컴퓨터情報學會論文誌 Vol.27 No.5

모델링 및 시뮬레이션은 목표 시스템의 동작 검증, 성능 분석, 운용 최적화, 예측을 위해 사용되는 기술이다. 이 기술의 대표적인 이산사건 시스템 명세(DEVS)는 모델들을 엄격한 형식론으로 정의하고 모델 간의 구조를 계층화한다. 이 DEVS 모델들의 원자 모델은 목표와 다른 의도로 동작하게 될 경우 시뮬레이션은 잘못된 의사결정으로 이어질 수 있다. 그럼에도 대부분 DEVS 시스템은 모델 테스트의 부재 또는 수동 테스트 환경으로 제공하여 개발자가 모델을 검증하는 데 오랜 시간이 소비된다. 본 논문에서는 파이썬 기반 DEVS에서 정확하고 빠른 원자 모델의 검증을 위해 스크립트 기반 테스트 시스템을 제안한다. 제안 테스트 시스템은 기존 방식인 수동 테스트와 새로운 방식인 스크립트 기반 테스트를 둘 다 사용한다. 우리 시스템의 실험 결과, 제안 테스트 방식은 스크립트를 10번 연속 실행 시 24ms 이내에 실행되었다. 그리하여 제안 시스템은 스크립트 기반 테스트를 사용해서 빠른 원자 모델 검증 시간을 보장하고, 테스트 스크립트의 재사용성을 향상한다. Modeling and simulation is a technique used for operational verification, performance analysis, operational optimization, and prediction of target systems. Discrete Event System Specification (DEVS) of this representative technology defines models with a strict formalism and stratifies the structures between the models. When the atomic DEVS models operate with an intention different the target system, the simulation may lead to erroneous decision-making. However, most DEVS systems have the exclusion of the model test or provision of the manual test, so developers spend a lot of time verifying the atomic models. In this paper, we propose a script-based automated test system for accurate and fast validation of atomic models in Python-based DEVS. The proposed system uses both the existing method of manual testing and the new method of the script-based testing. As Experimental results in our system, the script-based test method was executed within 24 millisecond when the script was executed 10 times consecutively. Thus, the proposed system guarantees a fast verification time of the atomic models in our script-based test and improves the reusability of the test script.

수정 IWAN 모델을 이용한 사질토의 반복경화거동에 대한 연구 2 : 모델 검증

이진선,김동수,추연욱,윤종구 한국지진공학회 2003 한국지진공학회논문집 Vol.7 No.5

본 논문에서는 동반된 논문에서 제안된 수정 병렬 IWAN 모델의 적용성을 금강 모래와 토요라 모래를 대상으로 비틂전단실험을 수행하여 검증하였다. 두가지 사질토에 대해서 대칭 하중과 불규칙 하중을 반복 재하하여 실험을 수행하였다. 제안된 모델의 변수는 다양한 상대밀도와 구속압 조건하에서 대칭 반복하중을 재하하여 결정하였다. 시험 결과, 하중반복회수가 증가함에 따라 발생하는 사질토의 반복경화거동을 수정 병렬 IWAN 모델로 표현 가능하였으며, Pyke(1979)에 의해 제안된 불규칙 하중 형태를 이용한 실험결과를 실험 조건이 유사한 대칭 반복하중 재하실험 결과로 얻어진 모델 변수를 이용하여 예측한 결과 실험 결과와 잘 일치함을 알 수 있었다. In order to verify the applicability of the developed modified parallel IWAN model. two types of cyclic torsional shear tests were performed using Kum-Kang and Toyoura sands. One was a symmetric-limit loading test and the other was an irregular loading test. Model parameters were derived from the symmetric limit loading tests at various relative densities and confining pressures. The modified parallel IWAN model can predict the cyclic hardening behavior of sands very well as increasing loading cycles in the symmetric-limit tests. Irregular loading tests were performed using the loading shape suggested by Pyke(1979). Cyclic behaviors under irregular loading were simulated using model parameters derived from symmetric limit loading test results of similar loading conditions. The predicted cyclic hardening behaviors under irregular loading matched well with experimental results and the applicability of the proposed model was verified.

MODEL-BASED AUTOMATIC TEST CASE GENERATION FOR AUTOMOTIVE EMBEDDED SOFTWARE TESTING

Ki-Wook Shin,Dong-Jin Lim 한국자동차공학회 2018 International journal of automotive technology Vol.19 No.1

We propose a method to automatically generate software and hardware test cases from a UML model developed through a model-based development process. Where languages such as source-code languages are used within the model, input and expected values for each test case are generated using a custom parser. As a next step, unit test cases are combined to generate integration test cases using a bottom-up approach. Then these cases are converted into hardware test cases for approval testing of embedded systems, using XQuery and hardware mapping tables. We demonstrate this process by applying it to the power window switch module of a Hyundai Santa Fe vehicle. Our approach provides an automatic testing procedure for embedded systems developed by model-based methods, and generates test cases efficiently using a recombination of signals. In conclusion, our proposed method could help reduce the resources needed for test case generation from software to hardware.

이미지 기반 AI 모델의 품질 고도화를 위한 이미지 증강 및 테스트 방법의 융복합 연구

서정민,이우진 한국전시산업융합연구원 2023 한국과학예술융합학회 Vol.41 No.4

This study was initiated to solve the difficulty of constructing a test set without bias for each image feature to verify the quality of an image-based artificial intelligence model. In previous studies, it was difficult to use because a large amount of labeling data was required for model learning. The purpose of this study is to improve the performance of artificial intelligence models by organizing learning and test datasets without bias and testing them by cluster without manual classification or labeling. Two experiments were conducted on the artificial intelligence model DC-VTON of the virtual try-on and the results were compared. First, the test set was divided into five clusters each using two clustering methods and the model was tested. In the two methods, the deviation of the model quality indicator SSIM values of the clusters was compared. The deviation of convolutional neural network-based clustering was 0.022 larger than that of pixel value-based clustering, resulting in more differences in performance by cluster. Second, two experiments were conducted to add re-learning data for each cluster to reach the target value of the model's quality performance index. Comparing the SSIM average of clusters that had low performance with after re-learning, the differential addition method by cluster increased by 3.2% and the equal addition method by cluster increased by 0.64%. In clusters with low quality, complementing the learning data resulted in improved performance improvement. Based on the results of these studies, it is expected that the quality of image-based artificial intelligence models can be improved by clustering images, which are unstructured data, according to characteristics, to refine test sets and enhance cluster-specific test results.

한국어 말하기 평가 개발 모형 및 시험타당도 검사 절차

김정태 한국언어문화교육학회 2008 언어와 문화 Vol.4 No.3

This study aims to introduce a test development model to develop a Korean speaking test and validation procedure on it. There is a little published work on Korean speaking test development and validation on them. Therefore, this study attempts to initiate a discussion about it. To measure the Korean speaking proficiency of foreign learners of Korean, language testers need to consider a appropriate development model as a initial step in validation procedure because validation process on a test includes validity evidence gathering from the beginning of test development to the end (Fulcher, 2005; Kim, 2006). Based on the test development models of Bachman and Palmer(1996), Davidson and Lynch (2002), and Kim (2006), the study shows the feasibility of adopting the ADDIE model from the field of instructional technology into Korean language test development. ADDIE stands for (need) analysis, design, development, implementation, and evaluation. In each phase, the study describes features to consider for the development of a Korean speaking test. In the needs analysis phase, language testers collect information on Korean language tests by reverse-engineering test items, where reverse-engineering means test specification writing for a test. In the design and development phases, language testers plan test development procedures, and design and develop test items, test sets, and its administration. In the implementation phase, a test is administered as pilot or a main test. Finally, in the evaluation phase, validation is carried out by using all the information gathered from the previous phases. The information includes validity arguments, test results and their quantitative analysis, and rater training procedures. The test development process is not linear, but also iterative and cyclical. (Pai Chai University)