• 한국정밀공학회지
• /
• 제15권12호
• /
• pp.97-105
• /
• 1998

The fiber oriented condition inside fiber reinforced composite material is a basic factor of mechanical properties of composite materials. It is very important to meausure the fiber orientation angle for the determination of molding conditions, mechanical characteristics, and the design of composite materials. In the work, the fiber orientation distribution of simulation figure plotted by PC is measured using image processing in order to examine the accuracy of intersection counting method. The fiber orientation function measured by intersection counting method using image processing is compared with the calculated fiber orientation function. The results show that the measured value of fiber orientation function using intersection counting method is lower than the calculated value, because the number of intersection between the scanning line and the fiber with smaller fiber aspect ratio is counted less than with larger fiber aspect ratio.

• Journal of the Korean Data and Information Science Society
• /
• 제14권4호
• /
• pp.929-937
• /
• 2003

Kernel type estimations of discontinuity point at an unknown location in regression function or its derivatives have been developed. It is known that the discontinuity point estimator based on $Gasser-M\ddot{u}ller$ regression estimator with a one-sided kernel function which has a zero value at the point 0 makes a poor asymptotic behavior. Further, the asymptotic variance of $Gasser-M\ddot{u}ller$ regression estimator in the random design case is 1.5 times larger that the one in the corresponding fixed design case, while those two are identical for the local polynomial regression estimator. Although $Gasser-M\ddot{u}ller$ regression estimator with a one-sided kernel function which has a non-zero value at the point 0 for the modification is used, computer simulation show that this phenomenon is also appeared in the discontinuity point estimation.

• 센서학회지
• /
• 제20권3호
• /
• pp.151-155
• /
• 2011

A variety of pattern recognition algorithms including neural networks may be applicable to the identification of odors. In this paper, an identification technique for an electronic odor sensing system applicable to wound state monitoring is presented. The performance of the radial basis function(RBF) network is highly dependent on the choice of centers and widths in basis function. For the fine tuning of centers and widths, those parameters are initialized by an ill-conditioned genetic fuzzy c-means algorithm, and the distribution of input patterns in the very first stage, the stochastic gradient(SG), is adapted. The adaptive RBF network with singular value decomposition(SVD), which provides additional adaptation capabilities to the RBF network, is used to process data from array-based gas sensors for early detection of wound infection in burn patients. The primary results indicate that infected patients can be distinguished from uninfected patients.

• 설비공학논문집
• /
• 제22권10호
• /
• pp.719-726
• /
• 2010

In this paper, the appropriate thermal storage time of an air-conditioning system with slab thermal storage was considered by use of summer measurement values. Two standards of heat extraction rate and criterion function were established as the standard that evaluates appropriateness. When heat extraction rate was a standard, zero hour and seven hours were obtained as appropriate thermal storage time, in the case of evaluation by energy consumption and running cost individually. Also, when criterion function was a standard, the difference between energy consumption and running cost was small, it was because the weight function to room air temperature deviation was much bigger than heat extraction rate.

• 대한전자공학회:학술대회논문집
• /
• 대한전자공학회 2002년도 ITC-CSCC -3
• /
• pp.1626-1629
• /
• 2002

This paper describes novel methods to construct fuzzy inference rules with gradient descent. The present methods have a constructive mechanism of the rule unit that is applicable in two parameters: the central value and the width of the membership function in the antecedent part. The first approach is to create the rule unit at the nearest position from the input space, for the central value of the membership function in the antecedent part. The second is to create the rule unit which has the minimum width, for the width of the membership function in the antecedent part. Experimental results are presented in order to show that the proposed methods are effective in difference on the inference error and the number of learning iterations.

• Structural Engineering and Mechanics
• /
• 제54권4호
• /
• pp.725-740
• /
• 2015

This study presents critical buckling load optimization of the axially graded layered uniform columns. In the first place, characteristic equations for the critical buckling loads for all boundary conditions are obtained using the transfer matrix method. Then, for each case, square of this equation is taken as a fitness function together with constraints. Due to explicitly unavailable objective function for the critical buckling loads as a function of segment length and volume fraction of the materials, especially for the column structures with higher segment numbers, initially, prescribed value is assumed for it and then the design variables satisfying constraints are searched using Differential Evolution (DE) optimization method coupled with eigen-value routine. For constraint handling, Exterior Penalty Function formulation is adapted to the optimization cycle. Different boundary conditions are considered. The results reveal that maximum increments in the critical buckling loads are attained about 20% for cantilevered and pinned-pinned end conditions and 18% for clamped-clamped case. Finally, the strongest column structure configurations will be determined. The scientific and statistical results confirmed efficiency, reliability and robustness of the Differential Evolution optimization method and it can be used in the similar problems which especially include transcendental functions.

• 한국정보통신학회:학술대회논문집
• /
• 한국정보통신학회 2021년도 춘계학술대회
• /
• pp.72-74
• /
• 2021

다양한 분야에서 현재 활용되고 있는 딥러닝 과정은 데이터 준비, 데이터 전처리, 모델 생성, 모델 학습, 모델 평가로 구성 된다. 이중 모델 학습 과정에서 손실함수는 모델이 학습하면서 출력한 값을 실제 값과 비교하여 그 차이를 출력하게 되고, 출력된 손실값을 기반으로 모델은 역전파 알고리즘을 통해 손실값이 감소하는 방향으로 가중치를 수정해가며 학습을 진행한다. 본 논문에서는 바이오마커 추출을 위한 딥러닝 모델에서 사용될 신경망 출력 값의 손실도를 측정하여 출력해주는 다양한 손실함수를 분석하고 실험을 통해 최적의 손실함수를 찾아내고자 한다.

• 한국수학교육학회지시리즈B:순수및응용수학
• /
• 제30권1호
• /
• pp.43-65
• /
• 2023

In this paper, by using the difference analogue of Nevanlinna's theory, the authors study the shared-value problem concerning two higher order difference operators of a transcendental entire function with finite order. The following conclusion is proved: Let f(z) be a finite order transcendental entire function such that λ(f - a(z)) < ρ(f), where a(z)(∈ S(f)) is an entire function and satisfies ρ(a(z)) < 1, and let 𝜂(∈ ℂ) be a constant such that ∆_𝜂ⁿ⁺¹ f(z) ≢ 0. If ∆_𝜂ⁿ⁺¹ f(z) and ∆_𝜂ⁿ f(z) share ∆_𝜂ⁿ a(z) CM, where ∆_𝜂ⁿ a(z) ∈ S ∆_𝜂ⁿ⁺¹ f(z), then f(z) has a specific expression f(z) = a(z) + Be^Az, where A and B are two non-zero constants and a(z) reduces to a constant.

• 전기학회논문지
• /
• 제67권1호
• /
• pp.114-123
• /
• 2018

This paper concerns a design and learning method of softmax function neural networks based on K-means clustering. The partial discharge data Information is preliminarily processed through simulation using an Epoxy Mica Coupling sensor and an internal Phase Resolved Partial Discharge Analysis algorithm. The obtained information is processed according to the characteristics of the pattern using a Motor Insulation Monitoring System program. At this time, the processed data are total 4 types that void discharge, corona discharge, surface discharge and slot discharge. The partial discharge data with high dimensional input variables are secondarily processed by principal component analysis method and reduced with keeping the characteristics of pattern as low dimensional input variables. And therefore, the pattern classifier processing speed exhibits improved effects. In addition, in the process of extracting the partial discharge data through the MIMS program, the magnitude of amplitude is divided into the maximum value and the average value, and two pattern characteristics are set and compared and analyzed. In the first half of the proposed partial discharge pattern classifier, the input and hidden layers are classified by using the K-means clustering method and the output of the hidden layer is obtained. In the latter part, the cross entropy error function is used for parameter learning between the hidden layer and the output layer. The final output layer is output as a normalized probability value between 0 and 1 using the softmax function. The advantage of using the softmax function is that it allows access and application of multiple class problems and stochastic interpretation. First of all, there is an advantage that one output value affects the remaining output value and its accompanying learning is accelerated. Also, to solve the overfitting problem, L2-normalization is applied. To prove the superiority of the proposed pattern classifier, we compare and analyze the classification rate with conventional radial basis function neural networks.

• 대한전자공학회:학술대회논문집
• /
• 대한전자공학회 2001년도 제14회 신호처리 합동 학술대회 논문집
• /
• pp.389-392
• /
• 2001

A motion vector sensitivity function for abrupt scene change detection is presented. Proposed function detects a scene change by a static uni-value, not using threshold comparion method in real time and compressed domain. All abrupt scene change was detected.