Earticle

This paper considers a six sigma project for reducing the cost copper of the cable materials in a electric wire company. The project follows a disciplined process of five macro phases: define, measure, analyze, improve, and control (DMAIC). A process map is used to identify process input variables. Three key process input variables are selected by using an input variables are selected by using an input variable evaluation table: large cable, plating, and a twisted pair. DOE is utilized for finding the optimal process conditions of the three key process input variables. The implementing result of this six sigma project is enable for reducing of the 2.8% copper materials.

The purpose of this study is to improve the accuracy of the Building Energy Analysis Program ECO2. Energy prediction through building energy simulation in the design phase is becoming very important for building maintenance effectively and optimal energy use. The ECO2 program continues to raise controversy over the accuracy of the evaluation due to its interpretation using simple calculation methods. In this study, ECO2 is compared with a globally recognized detailed analysis program Design builder with the resulting impact difference of each input variable to improve the accuracy of the sensitivity. Design builder is assumed to be an accurate analytical value, and modeling is carried out in accordance with the Energy Saving Design Standard. Sensitivity analysis, which analyzes the change in analytical result values by varying one input variable, is conducted in two programs, respectively. The Air infiltration rate and lighting density showed an error interval in which the results did not change linearly, but changed rapidly.

The Nakdong River Estuary is a vital ecological and water resource facing management challenges due to seawater intrusion and seasonal hydrological changes. The continuous opening of the estuarine bank since 2022 has increased the need for accurate salinity prediction to support ecosystem restoration and water-resource management. This study aims to improve the reliability and efficiency of salinity prediction models by optimizing the input variables using exploratory data analysis. Multidimensional observational data collected from January 2019 to June 2022 were analyzed to identify the statistical characteristics and seasonal variability of key factors. Principal component analysis was used to select critical variables, and a data-driven variable selection process was implemented. Experiments using a long short-term memory deep learning model showed that the optimized variable set improved the prediction performance by approximately 4.5% compared to models using all available variables. These results demonstrate that a data-centric approach enables the development of robust and interpretable prediction models that can be utilized by non-specialists in environmental hydraulics. The findings are expected to expand the practical application of artificial intelligence technologies in environmental management and contribute to the development of data-driven decision-support systems in estuarine environments.

In the worst case, a temporary ignition source (also known as transient combustibles) between two electrical panels can damage both panels. Mitigation strategies for electrical panel fires were previously developed using fire modeling and risk analysis. However, since they do not comply with deterministic fire protection requirements, it is necessary to analyze the boundary values at which combustibles may damage targets depending on various factors. In the present study, a sensitivity analysis of input variables related to the damage threshold of two electrical panels was performed for dimensionless geometry using a Fire Dynamics Simulator (FDS). A new methodology using a damage evaluation map was developed to assess the damage of the electrical panel. The input variables were the distance between the electrical panels, the vertical height of the fuel, the size of the fire, the wind speed and the wind direction. The heat flux was determined to increase as the vertical distance between the fuel and the panel decreased, and the largest heat flux was predicted when the vertical separation distance divided by one half flame length was 0.3-0.5. As the distance between the panels increases, the heat flux decreases according to the power law, and damage can be avoided when the distance between the fuel and the panel is twice the length of the panel. When the wind direction is east and south, to avoid damage to the electrical panel the distance must be increased by 1.5 times compared to no wind. The present scale model can be applied to any configuration where combustibles are located between two electrical panels, and can provide useful guidance for the design of redundant electrical panels.

An estimating procedure is introduced for kernel Poisson regression when the input variables consist of numerical and categorical variables, which is based on the penalized negative log-likelihood and the component-wise product of two different types of kernel functions. The proposed procedure provides the estimates of the mean function of the response variables, where the canonical parameter is linearly and/or nonlinearly related to the input variables. Experimental results are then presented which indicate the performance of the proposed kernel Poisson regression.

Spatio-temporal load forecasting (STLF) is a foundation for building the prediction-based power map, which could be a useful tool for the visualization and tendency assessment of urban energy application. Constructing one point-forecasting model for each electricity cell in the geographic space is possible; however, it is unadvisable and insufficient, considering the aggregation features of electricity cells and uncertainties in input variables. This paper presents a new STLF method, with a data-driven framework consisting of 3 subroutines: multi-level clustering of cells considering their aggregation features, load regression for each category of cells based on SLS-SVRNs (sparse least squares support vector regression networks), and interval forecasting of spatio-temporal load with sampled blind number. Take some area in Pudong, Shanghai as the region of study. Results of multi-level clustering show that electricity cells in the same category are clustered in geographic space to some extent, which reveals the spatial aggregation feature of cells. For cellular load regression, a comparison has been made with 3 other forecasting methods, indicating the higher accuracy of the proposed method in point-forecasting of spatio-temporal load. Furthermore, results of interval load forecasting demonstrate that the proposed prediction-interval construction method can effectively convey the uncertainties in input variables.

In this paper, we design an Fuzzy-Neural Networks(FNN) by means of divisions of fuzzy input space with multi-input variables. Fuzzy input space of Yamakawa's FNN is divided by each separated input variable, but that of the proposed FNN is divided by mutually combined input variables. The membership functions of the proposed FNN use both triangular and gaussian membership types. The parameters such as apexes of membership functions, learning rates, momentum coefficients, weighting value, and slope are adjusted using genetic algorithms. Also, an aggregate objective function(performance index) with weighting value is utilized to achieve a sound balance between approximation and generalization abilities of the model. To evaluate the performance of the proposed model, we use the data of sewage treatment process.

Determination of input variables for artificial neural network (ANN) depends entirely on the judgement of a modeller. As the number of input variables increases, the training time for the resulting ANN increases exponentially. Moreover, larger number of input variables does not guarantee better performance. In this research, we employ Genetic Algorithm for selecting proper input variables that yield the best performance in training the resulting ANN.

최근 급격히 증가되고 있는 도시화로 인해 불투수지역의 비율이 감소되고 있으며, 이상기후로 인한 집중호우 및 게릴라성 폭우가 빈번히 발생하고 있다. 이러한 이유로 내수침수 및 인명피해, 막대한 재산피해를 가져오고 있으며, 피해를 방지하기 위해 빗물펌프장 등이 설치되고 있다. 그러므로 도시홍수 분석 및 도시유역에서의 유출해석을 위해서는 빗물펌프장을 고려하여 분석하는 것이 필수적이지만, 유출모형 및 빗물펌프장 운영자료들이 불확실성을 가지고 있어 정확한 모형 해석 및 분석에 어려움이 있다. 따라서 본 연구에서는 자료의 품질평가에 사용되는 Pedigree matrix를 수정하여 수정 Pedigree matrix를 제안하고, 이를 이용한 빗물펌프장을 포함한 도시유역의 유출해석에서 불확실성을 야기할 것으로 예상되는 자료 및 매개변수를 선정하였다. 선정된 자료들을 대상으로 Pedigree matrix 점수와 불확실성 지수를 이용한 불확실성에 대한 기여도를 평가한 결과, 유역폭, 투수 및 불투수유역의 조도계수가 가장 불확실성이 높게 나타나 정확한 매개변수 추정이 필요하다는 결론을 내릴 수 있었다. 이 결과는 도시유역의 유출 해석 및 빗물펌프장 운영 시 중요 매개변수 선정에 도움을 주리라 판단된다.

Due to the rapidly increased urbanization and climate change, impervious area becomes smaller and torrential rainfall is occurred frequently. These caused severe urban flood with severe property damages and human lives. To cope with the urban flood, pumping stations with retention/detention ponds have been constructed. Therefore, the pumping stations have to be included in the urban flooding system or rainfall-runoff systems for the accurate analysis, the uncertainty incurred by the rainfall-runoff models or input data and/or parameters caused difficulties. In this study, modified Pedigree matrix is proposed based on the original Pedigree matrix and input data and parameters which might cause the uncertainty in the systems are selected. The contribution is analyzed with the modified Pedigree matrix and uncertainty index. As a result, basin width and roughness of pervious and impervious areas have the highest contribution to the uncertainty and the more accurate estimation is required. The proposed results could be applied in the pumping station operations in the urban area.

소프트웨어 테스트는 시스템의 신뢰도를 판단하는 중요한 작업이지만, 많은 노력과 비용을 필요로 한다. 모델 기반 테스트는 이러한 비용을 줄이기 위한 방안으로써 제안되었다. 정형적 모델로부터 시스템의 실행 가능한 경로를 파악하고, 각 경로마다 입력 값을 생성하여 테스트를 수행한다. 이 때, 적절한 입력 값을 찾기 위해 메타-휴리스틱 기법을 사용하는데, 기존의 알고리즘은 목적 경로와 관련이 없는 변수까지 구분없이 고려하기 때문에 시스템이 복잡할수록 불필요한 연산이 많아지는 문제가 있다. 본 논문은 슬라이싱 기법과 우선순위 정책을 적용한 테스트 데이터 자동 생성 기법을 제안하며, 실험을 통해 기존의 방법보다 효과적으로 테스트 데이터를 생성함을 보인다.

소프트웨어 테스트는 시스템의 신뢰도를 판단하는 중요한 작업이지만, 많은 노력과 비용을 필요로 한다. 모델 기반 테스트는 이러한 비용을 줄이기 위한 방안으로써 제안되었다. 정형적 모델로부터 시스템의 실행 가능한 경로를 파악하고, 각 경로마다 입력 값을 생성하여 테스트를 수행한다. 이 때, 적절한 입력 값을 찾기 위해 메타-휴리스틱 기법을 사용하는데, 기존의 알고리즘은 목적 경로와 관련이 없는 변수까지 구분없이 고려하기 때문에 시스템이 복잡할수록 불필요한 연산이 많아지는 문제가 있다. 본 논문은 슬라이싱 기법과 우선순위 정책을 적용한 테스트 데이터 자동 생성 기법을 제안하며, 실험을 통해 기존의 방법보다 효과적으로 테스트 데이터를 생성함을 보인다.

In the face of accelerating biodiversity loss and its significance in our coexistence with nature, biodiversity is becoming more crucial in sustainable development perspective. To estimate biodiversity in the future which provides valuable information for decision making system especially in the national level, a quantitative approach must be studied forehand as a baseline of the present status. In this study, we developed a large-scale map of Plant Species Richness (PSR, typical indicator of biodiversity) for Young-dong and Pyung-chang provinces. Due to the accessibility of appropriate data and advance of modelling techniques, reduction of variables without deteriorating the predictive power is considered by applying Genetic algorithm. In addition, a number of Correctly Classified Instances (CCI) with 10-fold cross validation which indicates the predictive power, was carried out for evaluation. This study, as a fundamental baseline, will be beneficial in future land work as well as ecosystem restoration business or other relevant decision making agenda.

인공신경망(artificial neural network) 기법은 인간의 두뇌 신경세포의 활동을 모형화한 것으로 오랜 시간동안 발전해 왔으며 여러 분야에서 활용되고 있고 수문분야에서도 인공신경망을 이용한 연구가 활발히 진행되어 왔다. RDAPS와 같은 단기수치예보 자료는 강우의 유무 판단과 같은 정성적인 분석에서 비교적 정확도가 높지만 정확한 강우량의 추정과 같은 정량적인 부분에서는 정확도가 매우 낮으므로 인공신경망 기법과 같은 후처리 기법을 통해서 정확도를 높이게 된다. 인공신경망 기법을 수행할 때, 가장 중요한 것은 입력변수선택(input variable selection)으로 입력 변수의 적절한 선택이 결과값에 큰 영향을 주게 된다. 본 연구에서는 mutual information을 입력 변수 선택 기법으로 채택하여, 인공신경망의 입력변수 선정의 정확도를 알아보고자 한다. Mutual information은 주어진 자료의 엔트로피값을 이용하여 변수들 간의 독립과 종속의 관계를 나타내는 기법으로서, MI값은 '0'에서 '1'의 값을 가지며 '0'에 가까울수록 변수들 간의 관계가 독립적이고 '1'에 가까울수록 종속적인 관계를 나타낸다. 인공신경망의 입력변수선정에 대한 mutual information의 정확도를 알아보기 위해, 기존 입력변수선택 기법과 mutual information을 이용했을 경우의 인공신경망의 처리능력, 정확도를 비교 검토하였다.

소프트웨어 테스트는 시스템의 신뢰도를 판단하는 중요한 작업이지만, 많은 노력과 비용이 요구된다. 모델 기반 테스트는 시스템 요구사항을 정형적으로 표현한 모델로부터 테스트 설계를 자동화함으로써 이러한 비용을 줄이기 위한 방안으로 제안되었다. 모델의 각 경로마다 입력값을 생성하여 테스트를 수행하는데, 이 때, 적절한 입력 값을 찾기 위해 메타-휴리스틱 기법을 사용한다. 본 논문은 슬라이싱 기법과 우선순위 정책을 적용한 테스트 데이터 자동 생성 기법을 제안하며, 목적 경로와 관련이 없는 변수를 제외하여 불필요한 계산을 억제한다. 실험을 통해 기존의 기법보다 효과적으로 테스트 데이터를 생성함을 보인다.

Software testing is important to determine the reliability of the system, a task that requires a lot of effort and cost. Model-based testing has been proposed as a way to reduce these costs by automating test designs from models that regularly represent system requirements. For each path of model to generate an input value to perform a test, meta-heuristic technique is used to find the test data. In this paper, we propose an automatic test data generation method using a slicing method and a priority policy, and suppress unnecessary computation by excluding variables not related to target path. And then, experimental results show that the proposed method generates test data more effectively than conventional method.

동일한 시험조건에서 반복시험으로부터 얻어진 실험 데이터는 이론적으로 동일한 값을 가져야 한다. 그러나 실제 데이터 결과는 다양한 환경 요소들에 의해 발생하는 오차와 불확실성을 가지게 되어 시험 값이 변동량을 가진다. 이는 정확한 실험 데이터를 얻는데 제한사항이 된다. 본 연구에서는 확률통계 방법을 이용하여 불확실성을 가진 입력변수의 유효범위를 결정하는 알고리즘을 제안하였다. 또한 실제 현장에서 사용되는 볼트 체결 마찰계수 데이터를 이용하여 제안된 알고리즘을 적용하여 불확실성을 내재한 입력변수의 유효범위를 산출하고 이에 대한 신뢰성 평가를 하였다.

Data points obtained by conducting repetitive experiments under identical environmental conditions are, theoretically, required to correspond. However, experimental data often display variations due to generated errors or noise resulting from various factors and inherent uncertainties. In this study, an algorithm aiming to determine valid bounds of input variables, representing uncertainties, was developed using probabilistic and statistical methods. Furthermore, a reliability assessment was performed to verify and validate applications of this algorithm using bolt-fastening friction coefficient data in a sample application.

For a friendly interaction between human and robot, emotional interchange has recently been more important. So many researchers who are investigating the emotion generation model tried to naturalize the robot's emotional state and to improve the usability of the model for the designer of the robot. And also the various emotion generation of the robot is needed to increase the believability of the robot. So in this paper we used the hybrid emotion generation architecture, and defined the generalized context input of emotion generation model for the designer to easily implement it to the robot. And we developed the personality and loyalty model based on the psychology for various emotion generation. Robot's personality is implemented with the emotional stability from Big-Five, and loyalty is made of familiarity generation, expression, and learning procedure which are based on the human-human social relationship such as balance theory and social exchange theory. We verify this emotion generation model by implementing it to the 'user calling and scheduling' scenario.

The study is aimed to develop fuzzy logic system that has overcurrent and saturation time as input variable and possibility of electrical fire as output variable by making bad conductor area with physical damage to indoor wiring. Most previous studies focused on thermal characteristics depending on the current size and no study considered the current size and saturation time at the same time. Therefore, the paper made into account current value and saturation time together. To this end, it created bad conductor area half the size of IV conductor (1.6 mm) on purpose and transmit electrical current from 10A to 60A by unit of 2A to find out the thermal characteristics and saturation time for current. Based on the data that came out, the study applied fuzzy logic and established the current and saturation time as input variable and chance of fire as output variable. As a result, the center of area of the system that depended only on the existing current value was 75 while the system that applied both current and saturation time presented the chance of fire at 92. It is found that the chance of bad conductor area and deteriorated insulation of electrical wire had current and saturation time as important variables. The data can be used as basic data like deteriorated wire insulation or operation features of circuit breaker in investigating the cause of electrical fire.

The prediction of algal bloom is an important field of study in algal bloom management, and chlorophyll-a concentration(Chl-a) is commonly used to represent the status of algal bloom. In, recent years advanced machine learning algorithms are increasingly used for the prediction of algal bloom. In this study, XGBoost(XGB), an ensemble machine learning algorithm, was used to develop a model to predict Chl-a in a reservoir. The daily observation of water quality data and climate data was used for the training and testing of the model. In the first step of the study, the input variables were clustered into two groups(low and high value groups) based on the observed value of water temperature(TEMP), total organic carbon concentration(TOC), total nitrogen concentration(TN) and total phosphorus concentration(TP). For each of the four water quality items, two XGB models were developed using only the data in each clustered group(Model 1). The results were compared to the prediction of an XGB model developed by using the entire data before clustering(Model 2). The model performance was evaluated using three indices including root mean squared error-observation standard deviation ratio(RSR). The model performance was improved using Model 1 for TEMP, TN, TP as the RSR of each model was 0.503, 0.477 and 0.493, respectively, while the RSR of Model 2 was 0.521. On the other hand, Model 2 shows better performance than Model 1 for TOC, where the RSR was 0.532. Explainable artificial intelligence(XAI) is an ongoing field of research in machine learning study. Shapley value analysis, a novel XAI algorithm, was also used for the quantitative interpretation of the XGB model performance developed in this study.