• Proceedings of the KIEE Conference
• /
• 1995.07b
• /
• pp.892-894
• /
• 1995

This paper deals with developmeant and application of coating thickness control system in hot dip galvanizing process. According to the line conditions, such as line speed, strip size and target coating weight, a predictive preset model sets the initial oprating conditions. Referring the zine coating informations from the gauge, mean coating value controller adjusts the chamber pressure and horizontal distance between strip and air knife, while coating deviation controller adjusts the lip gap profile of the air knife. All adaptive gains are interactively calculated by numeric models based on the theoretical analysis. The operating result with this system effectively reduces the coating deviation in transverse direction as well as in longitudinal direction.

• Earthquakes and Structures
• /
• v.17 no.1
• /
• pp.101-113
• /
• 2019

Earthquake-induced pounding is one of the major reasons for structural failure in earthquake prone cities. An accurate description of the pounding phenomenon of two buildings requires the consideration of systems with a large number of degrees of freedom including adequate contact impact formulations. In this paper, firstly, a node to surface formulation for the realization of state-of-the-art pounding models for structural beam elements is presented. Secondly, a hierarchical substructure technique is introduced, which is adapted to the structural pounding problem. The numerical accuracy and efficiency of the method, especially for the contact forces, are verified on an academic example, applying four different impact elements. Error estimations are carried out and compared with the classical modal truncation method. It is demonstrated that the hierarchical substructure method is indeed able to significantly speed up the numeric integration procedure by preserving a required level of accuracy.

• Nuclear Engineering and Technology
• /
• v.52 no.12
• /
• pp.2867-2873
• /
• 2020

In reprocessed uranium, derived from an impoverished fuel of light-water moderated reactors, there are isotopes of ^{232, 234, 236}U, which make its recycling remarkably difficult. A method of concentration of ²³⁵U target isotope in cascade's additional product was proposed to recover the isotopic composition of reprocessed uranium. A general calculation procedure is presented and a parameters' optimization of multi-flow cascades with additional products. For the first time a numeric model of a cascade that uses the cuts of partial flows of stages with relatively high separation factors was applied in this procedure. A novel computing experiment is carried out on separation of reprocessed uranium hexafluoride with providing a high concentration of ²³⁵U in cascade's additional product with subsequent dilution. The parameters of cascades' stages are determined so as to allow reducing the ^{232, 234, 236}U isotope content up to the acceptable. It was demonstrated that the dilution of selected products by the natural waste makes it possible to receive a low enriched uranium hexafluoride that meets the ASTM C996-15 specification for commercial grade.

• Journal of the Ergonomics Society of Korea
• /
• v.14 no.1
• /
• pp.97-104
• /
• 1995

The widespread use of VDU has improved the efficiency of information transmission between man and machine, but has caused new occupational health and ergonomics problems. In this study, we tried to construct a fuzzy hyman performance model of VDU workers in Korea. Fuzzy inferences of human perfor- mance are obtained from the fuzzy inference rule with the job difficulty, CFF, SACL, Type A. and the degree of concentration in VDU work. Eight healthy female undergraduate students at Kyungnam university for subjects aged 20 to 23 years were examined in this experiment. They calculated continuous addition, subtraction, and multiplication of 1 or 2 digit numbers that were produced randomly on the CRT. Subjects peoformed two types of a numeric operation, which easy and difficult work produced 400 and 600 problems within a 40 minute work session, respectively. Subjects were tested over two workdays according to the type of work(easy and difficult) consisting of four 40 minutes work sessions in the morning. Each work lasted for five minutes with a ten minutes rest break. 117 fuzzy inference rules were obtained from the experimental data. The value of consequent part was obtained by a descent method. The difference between real human error and estimated value of fuzzy inference was $1.8075{\pm}1.8591%(M{\pm}SD)$. The difference in easy and diffcult works were $2.69{\pm}2.13%$ and $0.92{\pm}0.93%$, respectively.

• Communications for Statistical Applications and Methods
• /
• v.27 no.3
• /
• pp.313-326
• /
• 2020

Deep Learning is the most important key to the development of Artificial Intelligence (AI). There are several distinguishable architectures of neural networks such as MLP, CNN, and RNN. Among them, we try to understand one of the main architectures called Recurrent Neural Network (RNN) that differs from other networks in handling sequential data, including time series and texts. As one of the main tasks recently in Natural Language Processing (NLP), we consider Neural Machine Translation (NMT) using RNNs. We also summarize fundamental structures of the recurrent networks, and some topics of representing natural words to reasonable numeric vectors. We organize topics to understand estimation procedures from representing input source sequences to predict target translated sequences. In addition, we apply multiple translation models with Gated Recurrent Unites (GRUs) in Keras on English-Korean sentences that contain about 26,000 pairwise sequences in total from two different corpora, colloquialism and news. We verified some crucial factors that influence the quality of training. We found that loss decreases with more recurrent dimensions and using bidirectional RNN in the encoder when dealing with short sequences. We also computed BLEU scores which are the main measures of the translation performance, and compared them with the score from Google Translate using the same test sentences. We sum up some difficulties when training a proper translation model as well as dealing with Korean language. The use of Keras in Python for overall tasks from processing raw texts to evaluating the translation model also allows us to include some useful functions and vocabulary libraries as well.

• Journal of KIISE:Software and Applications
• /
• v.30 no.5_6
• /
• pp.540-549
• /
• 2003

In collaborative recommendation two models are generally used: the user model and the article model. A user model learns correlation between users preferences and recommends an article based on other users preferences for the article. Similarly, an article model learns correlation between preferences for articles and recommends an article based on the target user's preference for other articles. In this paper, we investigates various combination methods of the user model and the article model for better recommendation performance. They include simple sequential and parallel methods, perceptron, multi-layer perceptron, fuzzy rules, and BKS. We adopt the multi-layer perceptron for training each of the user and article models. The multi-layer perceptron has several advantages over other methods such as the nearest neighbor method and the association rule method. It can learn weights between correlated items and it can handle easily both of symbolic and numeric data. The combined models outperform any of the basic models and our experiments show that the multi-layer perceptron is the most efficient combination method among them.

• The KIPS Transactions:PartD
• /
• v.18D no.1
• /
• pp.67-80
• /
• 2011

On an automated business document processing system maintaining financial data, errors on query based retrieval of numbers are critical to overall performance and usability of the system. Automatic spelling correction methods have been emerged and have played important role in development of information retrieval system. However scope of the methods was limited to the symbols, for example alphabetic letter strings, which can be reserved in the form of trainable templates or custom dictionary. On the other hand, numbers, a sequence of digits, are not the objects that can be reserved into a dictionary but a pure markov sequence. In this paper we proposed a new OCR model for spelling correction for numbers using the multiple streams and the context based correction on top of probabilistic information retrieval framework. We implemented the proposed error correction model as a sub-module and integrated into an existing automated invoice document processing system. We also presented the comparative test results that indicated significant enhancement of overall precision of the system by our model.

• Restorative Dentistry and Endodontics
• /
• v.19 no.2
• /
• pp.384-408
• /
• 1994

The purpose of this study was to evaluate the fracture resistance of tooth restored with gold inlay. A profound understanding of the isthmus width factor, which is one of the several parameters of cavity designs, would facilitate the appropriate cavity preparation in a specific clinical situation. In this study, the cavities for gold inlay were prepared in maxillary left first premolar. A three-dimensional model was designed using I-DEAS program. The model was composed of 2515- nodes and 2172 isoparametric brick elements. In the model isthmus width was varied into 1/4, 1/3 and 1/2 of intercuspal width respectively, and numeric values of the material properties of enamel, dentin and gold was set. Three types of load : concentrated load, divided load and distributed load was 500N. The empty cavities in the model were also examined using divided load and distributed load. The three - dimensional Finite Element Method was used to analysis the displacement and stress distribution. The results were as follows : 1. All of the experimental models which were filled with gold inlay revealed similar direction of displacement to that of the natural tooth model under the same load type. But in the models with empty cavities, as the isthmus width increased, the degree of displacement increased in the case of divided load type. 2. All experimental models which were filled with gold inlay showed stress concentration at load points, but in the models with empty cavities at divided load type, as isthmus width increased, stress was concentrated at the comer of the pulpal floor. 3. In the models with empty cavities at divided load type, tooth fracture was expected regardless of isthmus width, but all experimental models which were filled with gold inlay after cavity preparation were not susceptible to fracture. 4. In all experimental models which were filled with gold inlay after cavity preparation, displacement patterns were similar under both concentrated and divided load types. In the models with empty cavities, a divided load resulted in a bucco-lingual cuspal displacemenat in both sides, but a distributed load resulted in a lingual displacement of the tooth.

• The Transactions of the Korean Institute of Electrical Engineers D
• /
• v.53 no.6
• /
• pp.415-424
• /
• 2004

In this paper, we introduce and investigate a new category of rule-based fuzzy inference system based on Information Granulation(IG). The proposed rule-based fuzzy modeling implements system structure and parameter identification in the efficient form of “If..., then...” statements, and exploits the theory of system optimization and fuzzy implication rules. The form of the fuzzy rules comes with three types of fuzzy inferences: a simplified one that involves conclusions that are fixed numeric values, a linear one where the conclusion part is viewed as a linear function of inputs, and a regression polynomial one as the extended type of the linear one. By the nature of the rule-based fuzzy systems, these fuzzy models are geared toward capturing relationships between information granules. The form of the information granules themselves becomes an important design features of the fuzzy model. Information granulation with the aid of HCM(Hard C-Means) clustering algorithm hell)s determine the initial parameters of rule-based fuzzy model such as the initial apexes of the membership functions and the initial values of polynomial function being used in the Premise and consequence Part of the fuzzy rules. And then the initial Parameters are tuned (adjusted) effectively with the aid of the improved complex method(ICM) and the standard least square method(LSM). In the sequel, the ICM and LSM lead to fine-tuning of the parameters of premise membership functions and consequent polynomial functions in the rules of fuzzy model. An aggregate objective function with a weighting factor is proposed in order to achieve a balance between performance of the fuzzy model. Numerical examples are included to evaluate the performance of the proposed model. They are also contrasted with the performance of the fuzzy models existing in the literature.

• International conference on construction engineering and project management
• /
• 2015.10a
• /
• pp.401-404
• /
• 2015

Demand on free-form buildings is gradually increasing, yet owing to the difficulty of production-installation work, several problems occur in the construction phase upon construction of a building, including the increased cost and construction duration, and reduced constructibility. To solve these problems, a techonology to produce FCP using a CNC(Computerized Numeric Control) machine is developed. The technology is that the information of designed free-form buildings to the CNC machine is transferred, and the transferred information is used for RTM(Rod-Type Mold, the mold shaped by back-up rods) and PCM(Phase Change Material) shaping, and the shaped RTM and PCM have the role of molds to produce FCP. Construction duration and project cost are limited in building sites, so the efficiency of processes like production-installation of FCP for application of the technology is significant. Since it is almost impossible to change the production-installation process at the construction phase when they are established, process should be deliberately decided. Therefore, the purpose of the study is to propose a production-installation simulation model of free-form concrete panels, in aspect of PCM. This paper is establishing the process for production-installation of FCP, estimating time required by each construction type and proposing a time simulation model that changes according to various constraints based on the analyses. With the time simulation model, it will be possible to build a cost model and to review the optimal construction duration and project cost.