• 산업경영시스템학회지
• /
• 제40권4호
• /
• pp.203-210
• /
• 2017

Data clustering is one of the most difficult and challenging problems and can be formally considered as a particular kind of NP-hard grouping problems. The K-means algorithm is one of the most popular and widely used clustering method because it is easy to implement and very efficient. However, it has high possibility to trap in local optimum and high variation of solutions with different initials for the large data set. Therefore, we need study efficient computational intelligence method to find the global optimal solution in data clustering problem within limited computational time. The objective of this paper is to propose a combined artificial bee colony (CABC) with K-means for initialization and finalization to find optimal solution that is effective on data clustering optimization problem. The artificial bee colony (ABC) is an algorithm motivated by the intelligent behavior exhibited by honeybees when searching for food. The performance of ABC is better than or similar to other population-based algorithms with the added advantage of employing fewer control parameters. Our proposed CABC method is able to provide near optimal solution within reasonable time to balance the converged and diversified searches. In this paper, the experiment and analysis of clustering problems demonstrate that CABC is a competitive approach comparing to previous partitioning approaches in satisfactory results with respect to solution quality. We validate the performance of CABC using Iris, Wine, Glass, Vowel, and Cloud UCI machine learning repository datasets comparing to previous studies by experiment and analysis. Our proposed KABCK (K-means+ABC+K-means) is better than ABCK (ABC+K-means), KABC (K-means+ABC), ABC, and K-means in our simulations.

• 대한토목학회논문집
• /
• 제32권5D호
• /
• pp.527-532
• /
• 2012

지상라이다는 시간과 인력의 소모를 절감하면서도 고밀도의 포인트 클라우드를 생성할 수 있으므로 3차원 지적, 문화재보호, 건설관리 등에 활용되고 있다. 본 연구에서는 포인트 클라우드를 기반으로 한 실내 3차원 모델 구축 기법을 제시하였다. 제시한 방안은 세그먼테이션 단계와 외곽선 추출 단계로 이루어진다. 세그먼테이션 단계에서는 RANSAC과 정제격자를 이용하여 포인트 클라우드를 동일 평면에 따라 분할한다. 외곽선 추출단계에서는 외곽선 추출 격자를 이용하여 3차원 모델의 경계선을 추출한다. 또한 모델링에 사용되지 않은 포인트들을 메쉬자료화여 실내 공간의 세부 묘사를 위한 자료로 활용하였다. 제시한 모델링 기법은 메쉬자료를 이용하여 3차원 모델 구축을 하던 작업과정을 크게 개선하였다. 하지만 실내 공간의 환경 특성에 따라 RANSAC 경계값, 정제격자와 외곽선 추출 격자 크기의 조정이 필요하며 원형 또한 곡선 형태를 지닌 실내 구조물의 외곽선 추출을 위한 알고리즘의 개선이 필요하다.

• 한국환경과학회지
• /
• 제19권7호
• /
• pp.889-900
• /
• 2010

The experimental design and response surface methodology (RSM) have been applied to the investigation of the electro-UV complex process for the disinfection of E. coli in the water. The disinfection reactions of electro-UV process were mathematically described as a function of parameters power ($X_1$), NaCl dosage ($X_2$), initial pH ($X_3$) and disinfection time ($X_4$) being modeled by use of the Box-Behnken technique. The application of RSM using the Box-Behnken technique yielded the following regression equation, which is an empirical relationship between the residual E. coli number and test variables in actual variables: Ln (CFU) = 23.57 - 0.87 power - 1.87 NaCl dosage - 2.13 pH - 2.84 time - 0.09 power time - 0.07 NaCl dosage pH + 0.14 pH time + 0.03 $power^2$ + 0.47 NaCl $dosage^2$ + 0.20 $pH^2$+ 0.33 $time^2$. The model predictions agreed well with the experimentally observed result ($R^2$ = 0.9987). Graphical response surface and contour plots were used to locate the optimum point. The estimated ridge of maximum response and optimal conditions for the E. coli disinfection using canonical analysis was Ln 1.06 CFU (power, 15.40 W; NaCl dosage, 1.95 g/L, pH, 5.94 and time, 4.67 min). To confirm this optimum condition, the obtained number of the residual E. coli after three additional experiments were Ln 1.05, 1.10 and Ln 1.12. These values were within range of 0.62 (95% PI low)~1.50 (95% PI high), which indicated that conforming the reproducibility of the model.

• Imaging Science in Dentistry
• /
• 제48권2호
• /
• pp.103-109
• /
• 2018

Purpose: To analyze cone-beam computed tomography (CBCT) use, indications, and exposure parameters in San $Jos{\acute{e}}$, Costa Rica. Materials and Methods: A cross-sectional study was performed. All CBCT examinations over a period of 6 months at 2 radiological centers in San $Jos{\acute{e}}$, Costa Rica were evaluated. The examinations were performed with Veraview EPOC X550 and Veraviewepocs 3D R100 equipment. The patients' age and sex, clinical indication for CBCT, region of interest (ROI), repeat examinations, specialty of the referring dentist, field-of-view (FOV), tube voltage (kV), tube current (mA), and radiation dose (${\mu}Gy$) were evaluated. Patients were classified by age as children (${\leq}12years$), adolescents(13-18 years), and adults(${\geq}19years$). Results: The mean age of the 526 patients was 49.4 years. The main indications were implant dentistry and dental trauma. The most frequent ROIs were posterior, while anterior ROIs were much less common. The highest percentage of repeat examinations was in children. Fifty-six percent of the referring dentists were specialists. The most commonly used FOV was small. The mean tube voltage and current were 79.8 kV and 7.4 mA for Veraview EPOC X550 and 89.9 kV and 6 mA for Veraviewepocs 3D R100, respectively. The mean doses for children, adolescents, and adults were $6.9{\mu}Gy$, $8.4{\mu}Gy$, and $7.8{\mu}Gy$, respectively. Conclusion: Although CBCT was most commonly used in adults for implant dentistry, most repeat examinations were in children, and the highest mean dose was in adolescents. Additional dose optimization efforts should be made by introducing low-dose protocols for children and adolescents.

• 한국지능시스템학회논문지
• /
• 제19권6호
• /
• pp.749-754
• /
• 2009

4족 보행로봇의 빠른 걸음새를 자동으로 생성하는 문제에 대해서 GP(Genetic Programming)와 CPG(Central Pattern Generator) 기반의 두 가지 방식을 비교한다. GP(Genetic Programming)를 이용한 관절좌표계 상에서의 걸음새 생성 기법은 발끝의 자취와 수 많은 자세 파라미터를 사용하는 대신에 적은수의 관절 궤적을 생성하므로 효율적이다. CPG는 뇌로부터의 입력을 받아서 진동적인 출력을 생성하는 신경회로로 고등생물의 걸음 원리를 수학적으로 모델링한 것이다. 바이올로이드로 구성된 4족 보행로봇에 대하여 Webots기반의 ODE 시뮬레이션을 통해 접근 기법들에 대한 최적화를 수행하고 결과를 비교 분석한다. 그리고, 구해진 시뮬레이션과 결과를 실제 로봇에 대해서 각 동작을 실행시켜 보면서 CPG와 GP 기반의 걸음새 방식의 실제적인 성능 및 특성을 고찰한다.

• 드라이브 ㆍ 컨트롤
• /
• 제12권1호
• /
• pp.1-8
• /
• 2015

The stability of friction characteristics and thermal management for a dry type dual clutch transmission (DCT) are inferior to those of a wet clutch. Too high temperature resulting from frequent engagement of DCT speeds up degradation or serious wear of the pressure plate or burning of the clutch disk lining. Even though it is significantly important to estimate the temperature of a dry double clutch (DDC) in real-time, few meaningful study of the thermal model of DDC has been known yet. This study presented a thermal analytical model of lumped parameters for a DDC by analyzing its each component firstly. Then a series of experimental test was carried out on the test bench with a patented temperature telemetry system to validate the proposed thermal model. The thermal model, whose optimal parameter values were found by optimization algorithm, was also simulated on the experimental test conditions. The simulation results of DDC temperature show consistency with the experiment, which validates the proposed thermal model of DDC.

• 전기전자학회논문지
• /
• 제5권1호
• /
• pp.1-7
• /
• 2001

본 논문에서 역 방향 링크 채널에 대해 비 선형 등화기를 이용하여 CDMA 셀룰라 시스템을 연구하였다. 일반적으로 무선 통신에서 불확실한 채널 특성 때문에 Observable 들의 확률분포는 유한 세트의 파라미터로 규정될 수 없다. 대신에 training 샘플에 기반을 둔 Quantile과 Vector Quantizer를 사용함으로서 유한 수의 disjoint된 영역으로 m차 샘플 공간으로 분할하였다. 제안된 알고리듬은 RMSA 알고리즘에 의해 예측된 Quantile와 조건부 분할 모멘트에 따른 regression function의 부분적인 근사에 근간을 두고 있다. 본 논문의 등화기와 검출기는 잡음 분포의 Variation에 민감하지 않다는 관점에서 상당히 강한 특성을 보여 준다. 주요 아이디어는 Robust equalizer와 Robust partition detector가 어떤 환경의 무선 채널 하에서도 partition되지 않은 Observation space의 일반적인 등화기 보다 Observation의 등 확률로 분할된 부 공간에서 더 낳은 성능을 보여 준다. 또한 이런 개념을 CDMA 시스템에 적용하여 BER 성능을 분석하였다.

• 한국지반공학회논문집
• /
• 제32권11호
• /
• pp.97-108
• /
• 2016

달 탐사시 탐사 로버는 반드시 필요하며, 특히 월면토와 로버 휠의 상호작용에 의한 로버 휠의 성능은 로버의 최적 형상을 결정하는데 있어서 매우 중요하다. 본 연구에서는 한국형 인공 월면토(KLS-1)에서 달 탐사 로버 휠의 거동 성능을 평가하기 위하여 단일 휠 성능평가 실험장비를 개발하였고, 이를 이용하여 그라우져 유무에 따른 휠 성능 평가 실험을 수행하였다. 휠 성능은 슬립율에 따른 견인력, 토크, 침하 등으로 평가하였으며, 실험 결과 개발된 단일 휠 성능평가 실험장비는 휠 성능을 적절히 평가하는 것으로 나타났으며, 한국형 인공 월면토에서 그라우져가 있는 휠이 그라우져가 없는 휠에 비해 높은 견인 성능을 보여주었다. 향후 본 실험은 한국형 로버의 최적 휠 결정을 위해 사용될 수 있을 것으로 판단된다.

• 한국전기전자재료학회논문지
• /
• 제22권5호
• /
• pp.448-454
• /
• 2009

Hydrogen was produced from water plasma excited in high frequency (HF) inductively coupled tubular reactor. Mass spectrometry was used to monitor gas phase species at various process conditions, Water dissociation rate depend on the process parameters such as ICP power, $H_{2}O$ flow-rate and process pressure, Water dissociation percent in ICP reactor decrease with increase of chamber pressure, while increase with increase of ICP power and $H_{2}O$ flow rate. The effect of $CH_4$ gas addition to a water plasma on the hydrogen production has been studied in a HF ICP tubular reactor. The main roles of $CH_4$ additive gas in $H_{2}O$ plasma are to react with 0 radical for forming $CO_x$ and CHO and resulting additional $H_2$ production. Furthermore, $CH_4$ additives in $H_{2}O$ plasma is to suppress reverse-reaction by scavenging 0 radical. But, process optimization is needed because $CH_4$ addition has some negative effects such as cost increase and $CO_x$ emission.

• 전기학회논문지
• /
• 제59권2호
• /
• pp.436-444
• /
• 2010

In this study, we introduce and discuss a concept of a granular-oriented radial basis function neural networks (GRBF NNs). In contrast to the typical architectures encountered in radial basis function neural networks(RBF NNs), our main objective is to develop a design strategy of GRBF NNs as follows : (a) The architecture of the network is fully reflective of the structure encountered in the training data which are granulated with the aid of clustering techniques. More specifically, the output space is granulated with use of K-Means clustering while the information granules in the multidimensional input space are formed by using a so-called context-based Fuzzy C-Means which takes into account the structure being already formed in the output space, (b) The innovative development facet of the network involves a dynamic reduction of dimensionality of the input space in which the information granules are formed in the subspace of the overall input space which is formed by selecting a suitable subset of input variables so that the this subspace retains the structure of the entire space. As this search is of combinatorial character, we use the technique of genetic optimization to determine the optimal input subspaces. A series of numeric studies exploiting some nonlinear process data and a dataset coming from the machine learning repository provide a detailed insight into the nature of the algorithm and its parameters as well as offer some comparative analysis.