• 대한전자공학회:학술대회논문집
• /
• 대한전자공학회 2003년도 신호처리소사이어티 추계학술대회 논문집
• /
• pp.393-396
• /
• 2003

This paper proposes a method to speed up the document image binarization using a water flow model. The proposed method extracts the region of interest (ROI) around characters from a document image and restricts pouring water onto a 3-dimensional terrain surface of an image only within the ROI. The amount of water to be filled into a local valley is determined automatically depending on its depth and slope. Then, the proposed method accumulates weighted water not only on the locally lowest position but also on its neighbors. Finally, the depth of each pond is adaptively thresholded for robust character segmentation. Experimental results on real document images shows that the proposed method has attained good binarization performance as well as remarkably reduced processing time compared with that of the existing method based on a water flow model.

• 대한전기학회:학술대회논문집
• /
• 대한전기학회 1998년도 하계학술대회 논문집 B
• /
• pp.563-565
• /
• 1998

In this paper, we are proposed optimization method of fuzzy model in order to complex and nonlinear system. In the fuzzy modeling, a premise identification is very important to describe the charateristics of a given unknown system. Then, the proposed fuzzy model implements system structure and parameter identification, using the fuzzy inference method and genetic algorithms. Inference method for fuzzy model presented in our paper include the simplified inference and linear inference. Time series data for gas furance and sewage treatment process are used to evaluate the performance of the proposed model. Also, the performance index with weighted value is proposed to achieve a balance between the results of performance for the training and testing data.

• 제어로봇시스템학회:학술대회논문집
• /
• 제어로봇시스템학회 1993년도 한국자동제어학술회의논문집(국내학술편); Seoul National University, Seoul; 20-22 Oct. 1993
• /
• pp.233-238
• /
• 1993

In this paper, two diagnosis algorithms using the simple fuzzy, cognitive map (FCM) that is an useful qualitative model are proposed. The first basic algorithm is considered as a simple transition of Shiozaki's signed directed graph approach to FCM framework. And the second one is an extended version of the basic algorithm. In the extension, three important concepts, modified temporal associative memory (TAM) recall, temporal pattern matching algorithm and hierarchical decomposition are adopted. As the resultant diagnosis scheme takes short computation time, it can be used for on-line fault diagnosis of large scale and complex processes that conventional diagnosis methods cannot be applied. The diagnosis system can be trained by the basic algorithm and generates FCM model for every experienced process fault. In on-line application, the self-generated fault model FCM generates predicted pattern sequences, which are compared with observed pattern sequences to declare the origin of fault. In practical case, observed pattern sequences depend on transport time. So if predicted pattern sequences are different from observed ones, the time weighted FCM with transport delay can be used to generate predicted ones. The fault diagnosis procedure can be completed during the actual propagation since pattern sequences of tvo different faults do not coincide in general.

• 대한전기학회:학술대회논문집
• /
• 대한전기학회 1987년도 전기.전자공학 학술대회 논문집(II)
• /
• pp.1560-1563
• /
• 1987

This paper is concerned with time delay modeling of ED MOS gates which takes into account the slope of input waveform as well as the load condition. Defining the delay time as the time required to charge/discharge the load to the physical reference level, the rise/fall delay times arc derived in an explicit formula in terms of the sum of optimally weighted current unbalances at two end points of voltage transition. The proposed model is computationally effective and the error is typically within 10% of the SPICE results.

• 국제학술발표논문집
• /
• The 3th International Conference on Construction Engineering and Project Management
• /
• pp.1134-1139
• /
• 2009

Haul routes for earthmoving operation need to be carefully selected because the decision on the haul routes could make a significant difference in the project's cost and time. This paper proposes a decision support system for improving productivity of earthmoving operation that helps construction managers choose the best haul routes of trucks based on time slots. Also, a methodology for optimal selection of haul routes considering traffic conditions and topographic conditions of the routes is explained. Raster data model is used to find an available shortest path based on cost weighted distance. A system was developed on a geographic information system environment for efficient database management and easy manipulation of graphical data. A real-world case study was conducted to verify the applicability of the proposed system.

• 한국정보통신학회논문지
• /
• 제16권8호
• /
• pp.1593-1598
• /
• 2012

본 논문에서는 실시간으로 변하는 차량에 따라 퍼지 제어 기법을 이용하여 교통 신호를 효율적으로 제어하는 방법을 제안한다. 제안된 교통 신호 제어 방법은 교차로 각 차선에 대해 구간을 나누어서 각 구간의 차량의 수와 점유 시간에 대한 소속 함수를 설계한다. 설계된 소속 함수에서 소속도를 구한 후 퍼지 제어 규칙을 적용하여 Max-Min 방법으로 추론한다. 그리고 비퍼지화를 수행한 후에 우선순위를 정한다. 우선순위가 같을 경우에는 점유시간과 차량 수의 비율의 합, 각 구간에 대한 표준편차를 이용하여 우선순위를 정한다. 가장 높은 우선순위를 가지는 현시에 대해 진행 신호를 부여하고 나머지는 대기한다. 진행 신호의 시간은 교차로에 대기하는 모든 차량의 수와 진행 신호를 부여받은 현시의 차량의 수를 이용하여 계산한다. 본 논문에서 제안된 퍼지 신호 제어 기법과 정적인 신호 제어 기법에 대해 시뮬레이션을 통해 실험한 결과, 차량 혼잡도와 상관없이 제안된 차량 신호 제어 기법이 동일한 시간에 더 많은 차량이 원활하게 소통되는 것을 확인하였다.

• 대한기계학회논문집A
• /
• 제21권7호
• /
• pp.1030-1041
• /
• 1997

The control model in the tandem cold rolling mill consists of many mathematical theories and is used to calculate the reference values such as the roll gap and the rolling speed for good operation of rolling mill. But, the control model used presently has a problem causing inaccurate prediction of the rolling force. By the parameter identification, it was found that the main factor causing inaccurate prediction of the rolling force was incorrect modeling of the friction coefficient and the flow stress. To get rid of the erroneous factor new adaptive schemes are suggested in this work. Those are a long-time adaptation by the iterative least-square method and a short-time adaptation by the recursive weighted least-square method respectively. The new equations for the friction coefficient and the flow stress are derived by applying the suggested adaptive algorithms. Through the on-line test in an actual mill, it is proved that the rolling force predicted by the new equations is more accurate than the one by the existing equations ever used.

• 대한교통학회지
• /
• 제28권2호
• /
• pp.65-75
• /
• 2010

공공투자사업평가 시 사회적할인율의 높고 낮음에 따라 비용 편익분석의 결과가 상이하다는 점에서 적정 수치는 매우 중요한 요소이다. 그리고 사회적할인율의 개념을 무엇으로 적용하는 것이 합리적인지에 대한 많은 논란과 연구가 현재까지 진행되고 있다. 본 연구에서는 사회적할인율의 개념을 크게 세 가지 사회적 시간선호율, 사회적 투자수익률 그리고 이 둘을 가중평균 한 값으로 구분하였다. 추정방법으로는 사회적 시간선호율의 경우 소비이자율과 Pearce and Ulph식을 사용하였고, 사회적 투자수익률은 민간 총 자본 수익률을 사용하였다. 그리고 가중평균은 투자의 잠재가격을 이용한 Squire, L., Herman G. van der Tak식과 공공투자와 대치되는 민간투자와 민간 소비지출비율을 이용하여 추정하였다. 본 연구에서는 현재 시점의 사회적할인율을 추정하기 위해 연대구분을 1990년대, 2000년~2003년, 2004년~2008년으로 각각 구분한 후, 과거 한국개발연구원에서 제시된 1999년 7.5%, 2004년 6.5%, 그리고 2007년~현재 5.5%와 각각 비교하였다. 추정결과는 1990년대 6.6~10.7%, 2000년~2003년 4.0~7.0%, 2004년~2008년 2.4~3.9%로 각각 추정되었다. 현재 공공투자 사업평가에 사용되고 있는 사회적할인율 5.5%는 본 연구의 추정결과 보다 약1.6~3.1% 높은 것으로 나타났다. IMF이후 급격하게 하락한 소비이자율은 계속적으로 낮아지고 있고 앞으로도 지속될 것으로 예측하고 있다. 따라서 공공투자 사업의 목적인 현 세대부터 장래세대까지 모두를 고려한 사회적 자원의 재분배를 실현시키기 위해 선 현재의 사회적할인율은 점차 하향조정할 필요가 있다.

• 한국환경과학회지
• /
• 제19권8호
• /
• pp.971-981
• /
• 2010

The objectives of this study were to measure ambient total gaseous mercury (TGM) concentrations in Seoul, to analyze the characteristics of TGM concentration, and to identify of possible source areas for TGM using back-trajectory based hybrid receptor models like PSCF (Potential Source Contribution Function) and RTWC (Residence Time Weighted Concentration). Ambient TGM concentrations were measured at the roof of Graduate School of Public Health building in Seoul for a period of January to October 2004. Average TGM concentration was $3.43{\pm}1.17\;ng/m^3$. TGM had no notable pattern according to season and meteorological phenomena such as rainfall, Asian dust, relative humidity and so on. Hybrid receptor models incorporating backward trajectories including potential source contribution function (PSCF) and residence time weighted concentration (RTWC) were performed to identify source areas of TGM. Before hybrid receptor models were applied for TGM, we analysed sensitivities of starting height for HYSPLIT model and critical value for PSCF. According to result of sensitivity analysis, trajectories were calculated an arrival height of 1000 m was used at the receptor location and PSCF was applied using average concentration as criterion value for TGM. Using PSCF and RTWC, central and eastern Chinese industrial areas and the west coast of Korea were determined as important source areas. Statistical analysis between TGM and GEIA grided emission bolsters the evidence that these models could be effective tools to identify possible source area and source contribution.

• International Journal of Fuzzy Logic and Intelligent Systems
• /
• 제13권3호
• /
• pp.186-199
• /
• 2013

We pose pattern classification as a density estimation problem where we consider mixtures of generative models under partially labeled data setups. Unlike traditional approaches that estimate density everywhere in data space, we focus on the density along the decision boundary that can yield more discriminative models with superior classification performance. We extend our earlier work on the recursive estimation method for discriminative mixture models to semi-supervised learning setups where some of the data points lack class labels. Our model exploits the mixture structure in the functional gradient framework: it searches for the base mixture component model in a greedy fashion, maximizing the conditional class likelihoods for the labeled data and at the same time minimizing the uncertainty of class label prediction for unlabeled data points. The objective can be effectively imposed as individual mixture component learning on weighted data, hence our mixture learning typically becomes highly efficient for popular base generative models like Gaussians or hidden Markov models. Moreover, apart from the expectation-maximization algorithm, the proposed recursive estimation has several advantages including the lack of need for a pre-determined mixture order and robustness to the choice of initial parameters. We demonstrate the benefits of the proposed approach on a comprehensive set of evaluations consisting of diverse time-series classification problems in semi-supervised scenarios.