• Journal of the Korean Institute of Telematics and Electronics B
• /
• v.33B no.12
• /
• pp.95-105
• /
• 1996

This paper presents structure optimization of a feedforward neural netowrk controller using the genetic algorithm. It is important to design the neural network with minimum structure for fast response and learning. To minimize the structure of the feedforward neural network, a genralization of multilayer neural netowrks, the genetic algorithm uses binary coding for the structure and floating-point coding for weights. Local search with an on-line learnign algorithm enhances the search performance and reduce the time for global search of the genetic algorithm. The relative fitness defined as the multiplication of the error and node functions prevents from premature convergence. The feedforward neural controller of smaller size outperformed conventional multilayer perceptron network controller.

• Proceedings of the Korean Operations and Management Science Society Conference
• /
• 2006.05a
• /
• pp.305-312
• /
• 2006

K-means clustering algorithm 에서 주로 이루어지는 랜덤 초기화 (random initialization) 방법은 전역 최적화된 해(global minimum)를 찾아내기에 문제점을 지니고 있다. 즉, 여러 횟수의 알고리듬 반복(iteration)을 실행하더라도 전역 최적화된 해를 찾아내기가 매우 힘들며 주어진 자료의 크기(data size)가 큰 경우에 있어서 이는 거의 불가능하다. 본 논문은 이러한 문제점들을 극복하기 위한 방안으로, wavelet을 이용하여 최적의 초기 군집 중심점(initial clustering center)들을 선택하는 방법을 제시한다. 즉, 웨이블릿을 이용한 효과적인 초기화 (initialization)를 통해서 작은 알고리듬 반복 횟수만으로도 전역 최적화에 도달하는 초기화 방법을 기술한다. 이런 초기화 방법이 군집 알고리즘에 사용될 경우, 온라인상에서 실시간 이루어지는 군집 분석에 큰 도움이 된 수 있다.

• Proceedings of the Korean Society of Precision Engineering Conference
• /
• 2001.04a
• /
• pp.1068-1071
• /
• 2001

Recently, the minimum line width shows a tendancy to decrease and the multi-level to increase in semiconductor. Therefore, a planarization technique is needed, which chemical polishing(CMP) is considered as one of the most important process. CMP accomplishes a high polishing performance and a global planarization of high quality. But there are several defects in CMP such as microscratches, abrasive contaminations, and non-uniformity of polished wafer edges. Spin Etching can improve the defects of CMP. It uses abrasive-free chemical solution instead of slurry. Wafer rotates and chemical solution is simultaneously dispensed on a whole surface of the wafer. Thereby chemical reaction is occurred on the surface of wafer, material is removed. On this study, TEOS film is removed by CMP and Spin Etching, the results are estimated at a viewpoint of material removal rate(MRR) and within wafer non-uniformity(WIWNU).

• Transactions of the Korean Society of Mechanical Engineers A
• /
• v.24 no.5 s.176
• /
• pp.1193-1202
• /
• 2000

The paper describes the use of genetic algorithms (GA's) to the minimum weight design of stiffened composite panels for buckling constraints. The proposed design problem is characterized by mixture of continuous and discrete design variables corresponding to panel elements and stacking sequence of laminates, respectively. Design space is multimodal and non-convex, thereby introducing the need for global search strategies. Advanced strategies in GA's such as directed crossover, multistage search and separated crossover are adopted to improve search ability and to save computational resource requirements. The paper explores the effectiveness of genetic algorithms and their advanced strategies in designing stiffened composite panels under various uniaxial compressive load conditions and the linrlit on stacking sequence of laminates.

• Proceedings of the Korean Society of Precision Engineering Conference
• /
• 1997.10a
• /
• pp.935-938
• /
• 1997

Recently, the minimum line width shows a tendency to decrease and the multi-level increase in semiconductor. Therefore, a planarization technique is needed and chemical mechanical polishing(CMP) is considered as one of the most suitable process. CMP accomplishes a high polishing performance and a global planarization of high quality. But there are several defects in CMP such as micro-scratches, abrasive contaminations, and non-uniformity of polished wafer edges. Wet etching include of Spin-etching can improve he defects of CMP. It uses abrasive-free chemical solution instead of slurry. On this study, ILD(INterlayer-Dielectric) was removed by CMP and wet-etching methods in order to investigate the superiority of wet etching mechanism. In the thin film wafer, the results were evaluated at a viewpoint of material removal rate(MRR) and within wafer non-uniformity(WIWNU). And pattern step height was also compared for planarization characteristics of the patterned wafer.

• Journal of the Korean Institute of Telematics and Electronics A
• /
• v.33A no.6
• /
• pp.32-41
• /
• 1996

We propose the improved GMSK (gaussian-filtered minimum shift keying) quadrature modulator using the FIR(finite impulse response )filter whose coefficients are obtained form the differnce of phase response, and design its ASIC (applicaton specific integrated circuit) which can be used for GSM (global system for mobile communication) digital cellular system and DCS 1800 (digital cellular system at 1800MHz) personal communication system. Input data become quantized I and Q channel 10 bit signal through cosine and sine ROM mapping after being filtered by the FIR filter whose normalized bandwidth is 0.3 and designed by considering intersymbol interference as well as sampling ratio. These two signals become the GMSK modulated I and Q channel signal through DAC (digital-to-analog converter) and 7th order analog chebyshev LPF(low pass filter) respectively. The difference between the ideal analog signal and its digitized signal is analyzed in terms of sampling noise, quantization noise, truncation noise and coefficient noise. And the effect of the LPF following the DAC is considered. The ASIC design of the GMSK quadrature modulator is also confirmed by an experiment.

• Bulletin of the Korean Chemical Society
• /
• v.32 no.7
• /
• pp.2319-2324
• /
• 2011

Molecular structures and chemical properties of 1-aminoimidazole derivatives have been investigated at high levels of density functional theories. Heat of formation, density, explosive performances and impact sensitivities have been estimated at the global minimum of potential energy surface. As more nitro groups are introduced, the explosive performances of 1-aminoimidazole derivatives are enhanced, while the impact sensitivity becomes more sensitive. A two-dimensional plot between explosive performance and impact sensitivity has been utilized to comprehend the technical status of new explosive candidates. Based on locations in the two-dimensional plot, 1-aminodinitroimidzole isomers appears to have a potential to be good candidates for insensitive explosives, and 1-aminotrinitroimidazole may become a powerful explosive molecule whose behavior is quite close to HMX.

• Proceedings of the KIEE Conference
• /
• 1995.11a
• /
• pp.9-11
• /
• 1995

In this paper, tile squirrel case induction motors required multi-objective function are designed. As the objective function of the optimization program, we select the linear combination of loss and mass of motors by using weighting factors. Optimization process is performed by using the evolution strategy (ES). ES is the algorithm that can find the global minimum. To verify validity of the proposed method, a sample design is tried.

• Journal of the Korean Society of Manufacturing Process Engineers
• /
• v.7 no.4
• /
• pp.135-141
• /
• 2008

Independent suspension axle and final reduction gear for special-purpose vehicles such as a armored vehicles are almost imported in Germany etc. so, developing them is necessary to save cost. In severe condition (open fields, water surface driving, obstacle pass), special-purpose vehicles must work well. Drop box, axle and final reduction gear performed static analysis. We know that is possible weight reduction. The purpose of this paper is to find out the optimal shape of final reduction gear's case by means of response surface methodology. The response surface method is the statistical method which can be applied to the non-sensitivity based optimization. The response surface which is constructed by the least square method contains only the polynomial terms so that the global maximum and minimum points are easily obtained.

• KIEAE Journal
• /
• v.12 no.1
• /
• pp.119-125
• /
• 2012

The objective of this study is to evaluate the electric lighting energy consumption carried out by Daysim program. A comparison between the measurement and simulated exterior global horizontal illuminance shows differences about 10% and it is very similar to the measurement. The interior illuminance simulated by Daysim are 18.9% lower than the measurement and simulated lighting energy consumption is 10% lower than the measurement. Corrected annual lighting energy simulation results show that the best case is the combination of occupancy switch-off and dimming system with automatic controlled blinds (E-3). In case of no blinds, it occasionally derives the minimum lighting energy consumption but it causes the glare, so we need to be careful for choosing the control strategy. For the overcast sky, the lighting energy consumption is not changed significantly by control strategy while the lighting energy in the clear sky is changed noticeably. So we must know the right strategy for each case to control the electric lights and blinds.