• Proceedings of the KIEE Conference
• /
• 2008.10b
• /
• pp.247-248
• /
• 2008

본 논문에서는 TYPE-2 퍼지 추론 기반의 RBF 뉴럴 네트워크(TYPE-2 Radial Basis Function Neural Network, T2RBFNN)를 설계하고 PSO(Particle Swarm Optimization) 알고리즘을 이용하여 모델의 파라미터를 동정한다. 제안된 모델의 은닉층은 TYPE-2 가우시안 활성 함수로 구성되며, 출력층은 Interval set 형태의 연결가중치를 갖는다. 여기에서 규칙 전반부 활성함수의 중심 선택은 C-means 클러스터링 알고리즘을 이용하고, 규칙 후반부 Interval set 형태의 연결가중치 결정에는 경사 하강법(Gradient descent method)을 이용한 오류 역전파 알고리즘을 사용하여 학습한다. 또한, 최적의 모델을 설계하기 위한 학습율 및 활성함수의 활성화 영역 결정에는 입자 군집 최적화(PSO; Particle Swarm Optimization) 알고리즘으로 동조한다. 마지막으로, 제안된 모델의 평가를 위하여 모의 데이터 집합(Synthetic dadaset)을 적용하고 근사화 및 일반화 능력에 대하여 토의한다.

• Journal of The Korean Society of Agricultural Engineers
• /
• v.54 no.2
• /
• pp.67-75
• /
• 2012

Different from other natural hazards including floods, drought advances slowly and spreads widely, so that the preparedness is quite important and effective to mitigate the impacts from drought. Evaluation and forecast the status of drought for the present and future utilizing the meteorological scenario for agricultural drought can be useful to set a plan for agricultural drought mitigation in agriculture water resource management. In this study, drought climate scenario model on the basis of historical drought records for preparing agricultural drought mitigation was developed. To consider dependency and correlation between various climate variables, this model was utilized the historical climate pattern using reference year setting of four drought levels. The reference year for drought level was determined based on the frequency analysis result of monthly effective rainfall. On the basis of this model, drought climate scenarios at Suwon and Icheon station were set up and these scenarios were applied on the water balance simulation of reservoir water storage for Madun reservoir as well as the soil moisture model for Gosam reservoir watershed. The results showed that drought climate scenarios in this study could be more useful for long-term forecast of longer than 2~3 months period rather than short-term forecast of below one month.

• The Transactions of the Korean Institute of Electrical Engineers A
• /
• v.48 no.9
• /
• pp.1167-1174
• /
• 1999

In this paper, we present a new technique for the local decomposition of convex structuring elements for morphological image processing. Local decomposition of a structuring element consists of local structuring elements, in which each structuring element consists of a subset of origin pixel and its eight neighbors. Generally, local decomposition of a structuring element reduces the amount of computation required for morphological operations with the structuring element. A unique feature of our approach is the use of linear integer programming technique to determine optimal local decomposition that guarantees the minimal amount of computation. We defined a digital convex polygon, which, in turn, is defined as a convex structuring element, and formulated the necessary and sufficient conditions to decompose a digital convex polygon into a set of basis digital convex polygons. We used a set of linear equations to represent the relationships between the edges and the positions of the original convex polygon, and those of the basis convex polygons. Further. a cost function was used represent the total processing time required for computation of dilation/erosion with the structuring elements in a decomposition. Then integer linear programming was used to seek an optimal local decomposition, that satisfies the linear equations and simultaneously minimize the cost function.

• Proceedings of the IEEK Conference
• /
• 2002.07b
• /
• pp.1236-1239
• /
• 2002

When we see an object, we usually can say what it is easily even for the case where the object isn't shown in the frontal view. However, it is difficult to believe that all views of every object we have ever seen are fully memorized in our brain. Possibly, when an object is shown, we have some typical views of the object in our brain through our past experience and reconstruct the view to recognize what the presented object is. Non-negative Matrix Factorization (NMF) is one of the methods to extract the basis images from sample data set. The prominent feature of this method is that the reconstructed image is obtained by only additions of the basis images with suitable positive weights. So NMF can be seen more biologically plausible method than any other feature extraction methods such as Vector Quantization (VQ) and principal Component Analysis (PCA). In this paper, we adopt NMF to extract the aspect features from the set of images, which consists of various views of a given object. Some experiments are shown how much well NMF can extract the aspect features than any other methods such as VQ and PCA.

• Bulletin of the Korean Chemical Society
• /
• v.32 no.2
• /
• pp.673-680
• /
• 2011

The compound 3-(2-Mercaptopyridine)phthalonitrile has been synthesized and characterized by IR, UV-vis, and X-ray single-crystal determination. The molecular geometry from X-ray determination of the title compound in the ground state has been compared using the Hartree-Fock (HF) and density functional theory (DFT) with the 6-31G(d) basis set. The calculated results show that the DFT and HF can well reproduce the structure of the title compound. The energetic behavior of the title compound in solvent media was examined using the B3LYP method with the 6-31G(d) basis set by applying the Onsager and polarizable continuum model. Using the TD-DFT and TD-HF methods, electronic absorption spectra of the title compound have been predicted and good agreement with the TD-DFT method and the experimental determination was found. The predicted nonlinear optical properties of the title compound are much greater than those of urea. Besides, molecular electrostatic potential of the title compound were investigated by theoretical calculations. The thermodynamic properties of the compound at different temperatures have been calculated and corresponding relations between the properties and temperature have also been obtained.

• Journal of Korea Technical Association of The Pulp and Paper Industry
• /
• v.43 no.3
• /
• pp.1-10
• /
• 2011

Proportional.Integral.Derivative control schemes continue to provide the simplest and effective solutions to most of the control engineering applications today. However PID controller is poorly tuned in practice with most of the tuning done manually which is difficult and time consuming. This research comes up with a soft computing approach involving Genetic Algorithm, Evolutionary Programming, and Particle Swarm Optimization and Ant colony optimization. The proposed algorithm is used to tune the PID parameters and its performance has been compared with the conventional methods like Ziegler Nichols and Lambda method. The results obtained reflect that use of heuristic algorithm based controller improves the performance of process in terms of time domain specifications, set point tracking, and regulatory changes and also provides an optimum stability. This research addresses comparison of tuning of the PID controller using soft computing techniques on Machine Direction of basics weight control in pulp and paper industry. Compared to other conventional PID tuning methods, the result shows that better performance can be achieved with the soft computing based tuning method. The ability of the designed controller, in terms of tracking set point, is also compared and simulation results are shown.

• Proceeding of EDISON Challenge
• /
• 2017.03a
• /
• pp.110-117
• /
• 2017

Multiscale Simulation은 sub-nano scale의 전자 구조에서부터 macro scale의 multibody system에 이르기까지 다양한 시간/공간 스케일을 관통하는 시뮬레이션 기법이다. 즉, 전자수준에서의 변화로 인한 분자 전체의 구조 변화와 그에 따른 기능의 변화를 알 수 있는 simulation 방법으로 다양한 스케일에서 분자의 정보를 얻을 수 있다는 점에서 최근 중요하게 여겨지는 시뮬레이션 방법 중 하나이다. 따라서 본 연구에서는 몇 가지의 EDISON_CHEM 솔버들을 조합하여 Multiscale Simulation의 가능성을 제시하고자 한다. 또한, 세부적으로 양자계산 시에 어떤 이론을 선택하여 계산하면 더 정확한지, basis set 선택 시 발생하는 basis set superposition error(BSSE)로 인한 분자 수준의 물성의 오차는 어느정도 인지 알아보고자 했다. 본 연구에서는 비교적 간단하지만 온실 가스이자 에너지원으로 각광받고 있는 메탄을 대상으로 하였다. 다양한 시공간 스케일을 다루는 에디슨 솔버들 중에 양자 수준의 계산을 할 수 있는 솔버로는 "GAMESS"를 이용했고, 이 결과를 통해 분자 수준의 물성을 알아보기 위한 솔버로는 "사용자 지정 역장을 사용한 일반 분자동력학(general_MD)"과 "두가지 서로 다른 종류의 LJ입자에 대한 분자동력 시뮬레이션 프로그램(sejong_lj))"을 이용했다. 메탄의 상 전이 과정에 대한 연구 결과 Hartree fock (HF) self-consistent theory를 통해 얻은 force field 보다는 Second-order Møller-Plesset (MP2) perturbation theory로 얻은 force field가 더 정확한 상 전이 온도를 예측한다는 것을 메탄의 coarse-grained simulation을 통해 알 수 있었다. 또한, MP2 이론으로 구한 force field에서 BSSE를 보정해주면 실험적으로 구한 메탄의 상 전이 온도와 더 근사한 값의 시뮬레이션 결과를 얻을 수 있었다. 이를 통해 전자 간의 상호작용을 더 정교하게 계산하는 MP2 이론으로 force field를 구해서 BSSE를 보정해주면 계산의 결과가 정확해진다는 것을 알 수 있었으며 이것이 EDISON_CHEM의 솔버들로 가능하다는 것을 제시하였다.

• Journal of the Korean Institute of Intelligent Systems
• /
• v.10 no.3
• /
• pp.188-193
• /
• 2000

This paper presents an approach to obtain a reduced neuro-fuzzy model for a plant. The Neuro-Fuzzy Network are compose of the Radial Basis Function Networks with Gausis membership and learned by using temporal back propagation. The dependency in rough set theory is used to eliminate rules. Dependency between the condition membership value of each rule in a model and the output of the plant can allow us to see how much contribution the rule is to identify the plant. While the reduced model maintains the same performance as the original one, the selection algorithm can minimize its complexity and redundancy of the structure.

• ETRI Journal
• /
• v.29 no.1
• /
• pp.8-17
• /
• 2007

In this paper, we present a time domain combined field integral equation formulation (TD-CFIE) to analyze the transient electromagnetic response from dielectric objects. The solution method is based on the method of moments which involves separate spatial and temporal testing procedures. A set of the RWG functions is used for spatial expansion of the equivalent electric and magnetic current densities, and a combination of RWG and its orthogonal component is used for spatial testing. The time domain unknowns are approximated by a set of orthonormal basis functions derived from the Laguerre polynomials. These basis functions are also used for temporal testing. Use of this temporal expansion function characterizing the time variable makes it possible to handle the time derivative terms in the integral equation and decouples the space-time continuum in an analytic fashion. Numerical results computed by the proposed formulation are compared with the solutions of the frequency domain combined field integral equation.

• Journal of the Korean Chemical Society
• /
• v.40 no.5
• /
• pp.317-326
• /
• 1996

The frequency dependent longitudinal polarizabilities ${\alpha}zz(\omega)$ and the second hyper-polarizabilities ${\gamma}zzzz(\omega)$ of the linear polyenes, $C_4H_6\;to\;C_{30}H_{32}$, have been evaluated using the ab initio time-dependent coupled perturbed Hartree-Fock (TDCPHF) theory with the 6-31G basis set. The ratios of the dynamic properties to the static values have been examined to illustrate the relative dispersion effect and extrapolated to the infinite polymer limit. Also the effect of interchain interaction for linear and nonlinear optical properties has been investigated for $C_4H_6$ and the theoretical discussion has been described to overcome the limitation of ab initio TDHF method in the resonance region.