• 한국정보처리학회논문지
• /
• 제4권8호
• /
• pp.2070-2079
• /
• 1997

본 논문에서는 호 수락 제어 문제를 해결하기 위해 퍼지 논리 제어기의 장점과 신경망의 학습 능력을 이용한 ATM 망의 호 수락 제어 시스템을 제안하였다. ATM 망의 새로운 호는 현재 서비스 중인 호의 서비스 품질(QoS : quality of service)이 영향을 받지 않을 경우 망에 접속이 된다. 신경망 호 수락 제어 시스템은 입/출력 패턴의 학습으로 예측성 잇게 호 수락/거절을 하는 시스템이다. 본 논문의 퍼지 신경망 호 수락 제어 시스템에서는 학습 속도 개선을 위해 학습율과 모맨텀 상수에 퍼지 추론을 적용하였다. 이 시스템은 시뮬레이션을 통해 기존의 신경망 방법과 퍼지 신경망 방법에서의 학습 횟수 측정으로 제안 알고리즘의 우수성을 검증하였다. 시뮬레이션 결과 퍼지 학습 규칙에 근거한 퍼지 신경망 CAC(call admission control) 방식이 종래의 신경망 이론에 근거한 CAC 방식보다 학습 속도면에서 약 5배의 속도 향상이 있었다.

• Korea-Australia Rheology Journal
• /
• 제14권4호
• /
• pp.161-174
• /
• 2002

A new technique for numerical calculation of viscoelastic flow based on the combination of Neural Net-works (NN) and Brownian Dynamics simulation or Stochastic Simulation Technique (SST) is presented in this paper. This method uses a "universal approximator" based on neural network methodology in combination with the kinetic theory of polymeric liquid in which the stress is computed from the molecular configuration rather than from closed form constitutive equations. Thus the new method obviates not only the need for a rheological constitutive equation to describe the fluid (as in the original Calculation Of Non-Newtonian Flows: Finite Elements St Stochastic Simulation Techniques (CONNFFESSIT) idea) but also any kind of finite element-type discretisation of the domain and its boundary for numerical solution of the governing PDE's. As an illustration of the method, the time development of the planar Couette flow is studied for two molecular kinetic models with finite extensibility, namely the Finitely Extensible Nonlinear Elastic (FENE) and FENE-Peterlin (FENE-P) models.P) models.

• Nuclear Engineering and Technology
• /
• 제33권5호
• /
• pp.457-467
• /
• 2001

It is well known that the performance of a fuzzy neural network strongly depends on the input features selected for its training. In its applications to sensor signal estimation, there are a large number of input variables related with an output As the number of input variables increases, the training time of fuzzy neural networks required increases exponentially. Thus, it is essential to reduce the number of inputs to a fuzzy neural network and to select the optimum number of mutually independent inputs that are able to clearly define the input-output mapping. In this work, principal component analysis (PCA), genetic algorithms (CA) and probability theory are combined to select new important input features. A proposed feature selection method is applied to the signal estimation of the steam generator water level, the hot-leg flowrate, the pressurizer water level and the pressurizer pressure sensors in pressurized water reactors and compared with other input feature selection methods.

• 한국지반환경공학회 논문집
• /
• 제19권5호
• /
• pp.13-22
• /
• 2018

The number of researches on the mechanical properties of cemented sand and gravel (CSG) materials and the application of the CSG Dam has been increased. In order to explain the technical scheme of strength prediction model about the artificial neural network, we obtained the sample data by orthogonal test using the PVA (Polyvinyl alcohol) fiber, different amount of cementing materials and age, and established the efficient evaluation and prediction system. Combined with the analysis about the importance of influence factors, the prediction accuracy was above 95%. This provides the scientific theory for the further application of CSG, and will also be the foundation to apply the artificial neural network theory further in water conservancy project for the future.

• Advances in nano research
• /
• 제17권4호
• /
• pp.351-367
• /
• 2024

This research uses a numerical technique and a neural network process to investigate the stability management of non-uniform cylindrical constructions with varying sizes. The non-uniform or truncated conical shapes vary in axial length. This complicated geometry results in partial differential equations in the mathematical explanation of stability performance. Furthermore, material distributions vary in the radial direction in functionally graded materials such as metal and ceramic. The governing equations are obtained from beam theory using the energy technique and non-classical size-dependent theory, respectively. These equations are then solved using both a numerical and neural network methodology. This research can potentially be utilized in nanotechnology to build and evaluate size-dependent non-uniform cylindrical structures. As a consequence, it will help to develop sophisticated nanoscale materials and architectures.

• 제어로봇시스템학회논문지
• /
• 제6권6호
• /
• pp.447-453
• /
• 2000

Using the scattering theory of laser light, we analyze the particle sizing method. The scattered profile measured by the photodetector is sampled, scale conditioned by a 32 channel analog-to-digital converter, and is analyzed with the transform matrix from the light energy signals to the weights of the particle sizes. The particle size distribution is classified using the Hopfield neural network method as well as the conventional nonnegative least square method.

• 대한기계학회논문집A
• /
• 제30권5호
• /
• pp.520-527
• /
• 2006

The design of major industrial facilities for the prevention of fatigue failure is customarily done by defining a set of transients and performing a calculation of cumulative usage factor. However, sometimes, the inherent conservatism or lack of details as well as unanticipated transients in old plant may cause maintenance problems. Even though several famous on-line monitoring and diagnosis systems have been developed world-widely, in this paper, a new system fur fatigue monitoring and life evaluation of crane is proposed to reduce customizing effort and purchasing cost. With regard to the system, at first, comprehensive operating transient data has been acquired at critical locations of crane. The real-time data were classified, by using adaptive resonance theory that is one of typical artificial neural network, into representative stress groups. Then the each classified stress pattern was mapped to calculated cumulative usage factor in accordance with ASME procedure. Thereby, promising results were obtained fur the crane and it is believed that the developed system can be applicable to other major facilities extensively.

• 한국감성과학회:학술대회논문집
• /
• 한국감성과학회 2002년도 춘계학술대회 논문집
• /
• pp.195-200
• /
• 2002

ART (Adaptive Resonance Theory [1]) neural network and its variations perform non-hierarchical clustering by unsupervised learning. We propose a scheme "arboART" for hierarchical clustering by using several ART1.5-SSS networks. It classifies multidimensional vectors as a cluster tree, and finds features of clusters. The Basic idea of arboART is to use the prototype formed in an ART network as an input to other ART network that has looser distance criteria (Ishihara, et al., [2,3]). By sending prototype vectors made by ART to one after another, many small categories are combined into larger and more generalized categories. We can draw a dendrogram using classification records of sample and categories. We have confirmed its ability using standard test data commonly used in pattern recognition community. The clustering result is better than traditional computing methods, on separation of outliers, smaller error (diameter) of clusters and causes no chaining. This methodology is applied to Kansei evaluation experiment data analysis.

• Journal of Information Processing Systems
• /
• 제16권5호
• /
• pp.1145-1157
• /
• 2020

A convolution neural networks (CNNs) has demonstrated outstanding performance compared to other algorithms in the field of face recognition. Regarding the over-fitting problem of CNN, researchers have proposed a residual network to ease the training for recognition accuracy improvement. In this study, a novel face recognition model based on game theory for call-over in the classroom was proposed. In the proposed scheme, an image with multiple faces was used as input, and the residual network identified each face with a confidence score to form a list of student identities. Face tracking of the same identity or low confidence were determined to be the optimisation objective, with the game participants set formed from the student identity list. Game theory optimises the authentication strategy according to the confidence value and identity set to improve recognition accuracy. We observed that there exists an optimal mapping relation between face and identity to avoid multiple faces associated with one identity in the proposed scheme and that the proposed game-based scheme can reduce the error rate, as compared to the existing schemes with deeper neural network.

• 한국지능시스템학회논문지
• /
• 제5권4호
• /
• pp.41-55
• /
• 1995

본 논문에서는 복잡한 비선형 시스템의 모델링을 위해 퍼지-뉴럴 네트워크(FNNs)를 사용한 최적 동적 방법이 제안된다. 제안된 퍼지-뉴럴 모델링은 공정시스템의입축력 데이타를 이용하여 기존의 최적이론, 언어적 퍼지구현규칙, 뉴럴네트워크 등의 지능형 이론을 도입하여 시스템의 구조와 파라미터 동정을 구현한다. 이 모델링의 추론형태는 간략추론이 사용된다. 최적 모델을 얻기위해, 퍼지-뉴렬 네트워크의 학습률과 모멘텀 계수가 본논문에서 제안한 개선된 컴플렉스 법과 수정된 학습알고리즘을 이용하여 자동동조 된다. 이 알고리즘의 비선형 공정으로의 응용을 위하여 교통 경로 선택 데이타 및 하수처리시스템의 활성화와 공정 데이타가 제안한 모델링의 성능을 평가하기 위해 사용된다. 제안된 방법이 기존의 다른 논문과 비교하여 더 높은 정확도를 가진 지능형 모델을 생성함을 보인다.