• Journal of the Korean Society for Industrial and Applied Mathematics
• /
• 제24권2호
• /
• pp.143-159
• /
• 2020

We propose a consistent numerical method for elliptic interface problems with nonhomogeneous jumps. We modify the discontinuous bubble immersed finite element method (DB-IFEM) introduced in (Chang et al. 2011), by adding a consistency term to the bilinear form. We prove optimal error estimates in L² and energy like norm for this new scheme. One of the important technique in this proof is the Bramble-Hilbert type of interpolation error estimate for discontinuous functions. We believe this is a first time to deal with interpolation error estimate for discontinuous functions. Numerical examples with various interfaces are provided. We observe optimal convergence rates for all the examples, while the performance of early DB-IFEM deteriorates for some examples. Thus, the modification of the bilinear form is meaningful to enhance the performance.

• Journal of Magnetics
• /
• 제18권2호
• /
• pp.202-206
• /
• 2013

In this paper, the MFL (magnetic flux leakage) signals of complex defects on the underground gas pipeline are analyzed and their depths are estimated. Since closely-located defects (complex defects) affect each other, accelerate the progress of defection, and are finally combined to one (cluster), it's meaningful to differentiate complex defects from single defects by analyzing their characteristics. Various types of complex defects are characterized and analyzed by defining the safety distance for interference. 26 artificial defects are carved on the pipeline simulation facility (PSF) to analyze the characteristics of complex defect and demonstrate the accuracy of the proposed complex defect estimation. The proposed method shows average length error of 5.8 mm, average width error of 15.55 mm, and average depth error of 8.59%, respectively.

• Nuclear Engineering and Technology
• /
• 제51권3호
• /
• pp.709-718
• /
• 2019

The international nuclear industry has undergone a lot of changes since the Fukushima, Chernobyl and TMI nuclear power plant accidents. However, there are still large and small component deficiencies at nuclear power plants in the world. There are many causes of electrical equipment defects. There are also factors that cause component failures due to human errors. This paper analyzed the root causes of failure and types of human error in 300 cases of electrical component failures. We analyzed the operating experience of electrical components by methods of root causes in K-HPES (Korean-version of Human Performance Enhancement System) and by methods of human error types in HuRAM+ (Human error-Related event root cause Analysis Method Plus). As a result of analysis, the most electrical component failures appeared as circuit breakers and emergency generators. The major causes of failure showed deterioration and contact failure of electrical components by human error of operations management. The causes of direct failure were due to aged components. Types of human error affecting the causes of electrical equipment failure are as follows. The human error type group I showed that errors of commission (EOC) were 97%, the human error type group II showed that slip/lapse errors were 74%, and the human error type group III showed that latent errors were 95%. This paper is meaningful in that we have approached the causes of electrical equipment failures from a comprehensive human error perspective and found a countermeasure against the root cause. This study will help human performance enhancement in nuclear power plants. However, this paper has done a lot of research on improving human performance in the maintenance field rather than in the design and construction stages. In the future, continuous research on types of human error and prevention measures in the design and construction sector will be required.

• 전자공학회논문지B
• /
• 제31B권7호
• /
• pp.170-178
• /
• 1994

The nearest neighbor classification rule is widely used because it is not only simple but the error rate is asymptotically less than twice Bayes theoretical minimum error. But the method basically use the whole training patterns as the reference vectors. so that both storage and classification time increase as the number of training patterns increases. LVQ(Learning Vector Quantization) resolved this problem by training the reference vectors instead of just storing the whole training patterns. But it is a heuristic algorithm which has no theoretic background there is no terminating condition and it requires a lot of iterations to get to meaningful result. This paper is to propose a new training method of the reference vectors. which minimize the given error function. The nearest neighbor network,the network version of the nearest neighbor classification rule is proposed. The network is funtionally identical to the nearest neighbor classification rule is proposed. The network is funtionally identical to the nearest neighbor classification rule and the reference vectors are represented by the weights between the nodes. The network is trained to minimize the error function with respect to the weights by the steepest descent method. The learning algorithm is derived and it is shown that the proposed method can adjust more reference vectors than LVQ in each iteration. Experiment showed that the proposed method requires less iterations and the error rate is smaller than that of LVQ2.

• 대한전자공학회논문지
• /
• 제26권2호
• /
• pp.87-93
• /
• 1989

패턴인식과 영상처리 응용에 이용되는 자승오차 패턴 클러스터링 알고리듬은 특징벡터 행렬의 연산에 상당한 처리시간은 요구한다. 그러므로 본 논문은 병렬처리와 파이프라인 특성을 갖는 잉여수체계를 이용한 고속의 자승오차 패턴 클러스터링 프로세서를 제안한다. 제안된 자승오차 패턴 클러스터링 프로세서는 영상분할 실험으로부터 의미있는 영역으로 나눌 수 있는 클러스터의 수에 대하여 만족할 만한 오차를 보이며 80287 수치 연산용 프로세서보다 약 200배 빠름을 보인다. 그 결과 대규모의 데이타를 실시간으로 처리하여야 하는 응용분야에 효과적으로 이용할 수 있음을 확인하였다.

• 한국지능시스템학회논문지
• /
• 제14권1호
• /
• pp.28-32
• /
• 2004

현재 구현되어 있는 한국어 철자 교정 시스템은 문장의 문법 정보나 연어 관계로부터 문장의 오류를 처리하는 방식을 쓰고 있다. 본 논문에서는. 홑문장에서 의미소 사이의 관계를 이용하여 오타 문자를 교정하고, 오타에 의한 의미적인 오류가 있을 때에는 적절한 의미를 가지는 단어로 대체하는 시스템을 제안한다. 상기의 제안된 시스템을 이용하여 의미소들 간의 의미가 통하는 여러 개의 문장들이 제공된다. 단어의 뜻에 따라 체언은 의미 트리를 형성하고, 서술어는 주어 및 목적어의 체언과 의미 관계를 정의한다. 오류가 포함된 문장에서, 의미 관계를 비교, 분석하여 주어 및 목적어의 체언이 틀렸을 경우에는 서술어로부터, 서술어가 틀렸을 경우에는 주어 및 목적어의 체언으로부터, 수식어가 틀렸을 경우에는 체언 또는 서술어로부터 정의된 상호 의미 관계를 이용하여 한 문자에 대한 오타를 교정하고 오타에 의한 의미적 오류가 발견될 때에는 상기와 같은 철자 교정 방법을 적용하였다.

• The Journal of Asian Finance, Economics and Business
• /
• 제6권4호
• /
• pp.83-90
• /
• 2019

The purpose of this study is to test the effect of business strategy on valuation error. Business strategy includes managerial decision making and managerial tendency. In previous research, there is a negative relationship between business strategy and accounting quality. In this study, we try to confirm whether strategy tendencies affected valuation errors. In order to confirm empirically between business strategy and overvaluation, we use 8,117 firms that between 2006 and 2017 and listed in KSE and KOSDAQ. We calculated business strategy which is introduced by Bentley, Omer, and Sharp (2013). We also used the overvaluation method introduced in Rhodes-Kropf, Robinson, and Viswanathan (2005). The results show that the more the leading business strategy is, the greater the value error becomes. In the case of dividing into leading and defensive companies, the lead firms showed a significant positive correlation with the valuation errors, while the defensive firms showed the negative relationship with overvaluation. This study examined the business strategy and the overvaluation. we confirmed whether the management strategy deepens the evaluation error caused by the firm characteristics. The results are meaningful that we extended the study on the quality of financial reporting of leading strategic firms.

• 한국환경과학회지
• /
• 제28권1호
• /
• pp.125-135
• /
• 2019

For the purposes of enhancing usability of Numerical Weather Prediction (NWP), the quantitative precipitation prediction scheme by machine learning has been proposed. In this study, heavy rainfall was corrected for by utilizing rainfall predictors from LENS and Radar from 2017 to 2018, as well as machine learning tools LightGBM and XGBoost. The results were analyzed using Mean Absolute Error (MAE), Normalized Peak Error (NPE), and Peak Timing Error (PTE) for rainfall corrected through machine learning. Machine learning results (i.e. using LightGBM and XGBoost) showed improvements in the overall correction of rainfall and maximum rainfall compared to LENS. For example, the MAE of case 5 was found to be 24.252 using LENS, 11.564 using LightGBM, and 11.693 using XGBoost, showing excellent error improvement in machine learning results. This rainfall correction technique can provide hydrologically meaningful rainfall information such as predictions of flooding. Future research on the interpretation of various hydrologic processes using machine learning is necessary.

• 대한수학회논문집
• /
• 제16권3호
• /
• pp.487-508
• /
• 2001

We consider the Cauchy problem for Laplacian. Using the single layer representation, we obtain an equivalent system of boundary integral equations. We show the singular values of the ill-posed Cauchy operator decay exponentially, which means that a small error is exponentially amplified in the solution of the Cauchy problem. We show the decaying rate is dependent on the geometry of he domain, which provides the information on the choice of numerically meaningful modes. We suggest a pseudo-inverse regularization method based on singular value decomposition and present various numerical simulations.

• 한국CDE학회논문집
• /
• 제21권2호
• /
• pp.151-158
• /
• 2016

It is important to analyze dimensional errors occurring during shipbuilding process. A ship is constructed by assembling blocks and installing outfits in assembled ship structure. Blocks and outfits have a main direction that has greater importance than other directions from the view point of dimensional error. Therefore, a main direction should have a greater weighting factor than other directions in order to achieve meaningful inspection results. In this paper, a modified point registration method based on iterative closest point (ICP) is proposed. In this method, a user determines one or two main directions among x, y, and z directions, and then each main direction is made to have a greater weighting factor than other directions. For points registration, mapping between measured points and design points are performed by the modified ICP in which weighting factor assigned to each main direction is considered.