• 한국정보통신학회논문지
• /
• 제16권3호
• /
• pp.448-459
• /
• 2012

신경회로망의 동작과 정보처리 능력 등에 관하여 살펴보고자 할 때, 신경회로망의 구성 요소를 어떻게 모델화 할 것인가는 중요한 문제이다. 소자의 응답특성이 바뀜에 따른 특성의 변화, 결합강도 및 적응규칙이 바뀜으로써 회로망 전체의 다이나믹스가 바뀌는 모습, 소자 상호간의 결합 형태에 따른 정보처리 능력의 변화 등과 같은 신경회로망이 가진 다양한 정보처리 능력을 밝히는 것은 병렬 정보처리의 메카니즘을 이해하는 문제와도 일맥상통하고 있다. 따라서 이러한 문제들에 대하여 신경회로망의 정보처리 능력을 해석적으로 평가하는 것은 병렬분산 정보처리의 본질을 밝힌다는 측면에서 중요하게 여겨진다. 따라서 본 논문에서는 신경회로망을 구성하는 구성요소의 변화, 그 가운데에서도 특히 소자의 히스테리시스 특성이 신경망의 계열연상능력에 미치는 영향에 대한 이론적 해석결과에 대하여 기술한다.

• 한국전산구조공학회:학술대회논문집
• /
• 한국전산구조공학회 2007년도 정기 학술대회 논문집
• /
• pp.49-54
• /
• 2007

Structural integrity assessment technique for pavement system is studied considering the uncertainties among the material properties. The artificial neural networks technique is applied for the inverse analysis to estimate the elastic modulus based on the measured deflections from the FWD test. A computer code based on the spectral element method was developed for the accurate and fast analysis of the multi-layered soil structures, and the developed program was used for generating the training and testing patterns for the neural network. Neural networks was applied to estimate the elastic modulus of pavement system using the maximum deflections with and without the uncertainties in the material properties. It was found that the estimation results by the conventiona1 neural networks were very poor when there exist the uncertainties and the estimation results could be significantly improved by adopting the proposed method for generating training patterns considering the uncertainties among material properties.

• International Journal of Fuzzy Logic and Intelligent Systems
• /
• 제6권1호
• /
• pp.33-38
• /
• 2006

In this paper, we propose a new architecture of Fuzzy Polynomial Neural Networks(FPNN) by means of genetically optimized Fuzzy Polynomial Neuron(FPN) and discuss its comprehensive design methodology involving mechanisms of genetic optimization, especially Genetic Algorithms(GAs). The conventional FPNN developed so far are based on mechanisms of self-organization and evolutionary optimization. The design of the network exploits the extended Group Method of Data Handling(GMDH) with some essential parameters of the network being provided by the designer and kept fixed throughout the overall development process. This restriction may hamper a possibility of producing an optimal architecture of the model. The proposed FPNN gives rise to a structurally optimized network and comes with a substantial level of flexibility in comparison to the one we encounter in conventional FPNNs. It is shown that the proposed advanced genetic algorithms based Fuzzy Polynomial Neural Networks is more useful and effective than the existing models for nonlinear process. We experimented with Medical Imaging System(MIS) dataset to evaluate the performance of the proposed model.

• 한국농공학회논문집
• /
• 제54권5호
• /
• pp.43-47
• /
• 2012

A evaluation for the strength of rock includes a lot of uncertainty due to existence of discontinuity surface and weakness plain in the rock mass, so essential test results and other data for the resonable strength analysis are absolutely insufficient. Therefore, a analytical technique to reduce such uncertainty can be required. A probabilistic analysis technique has mainly to make up for the uncertainty to investigate the strength of rock mass. Recently, a artificial neural networks, as a more newly analysis method to solve several problems in the existing analysis methodology, trends to apply to study on the rock strength. In this study the unconfined compressive strength from basic physical property values of sedimentary rock, black shale and red shale, distributed in Daegu metropolitan area is estimated, using the artificial neural networks. And the applicability of the analysis method is investigated. From the results, it is confirmed that the unconfined compressive strength of the sedimentary rock can be easily and efficiently predicted by the analysis technique with the artificial neural networks.

• Computers and Concrete
• /
• 제24권4호
• /
• pp.329-345
• /
• 2019

Despite the extensive use of mortar materials in constructions over the last decades, there is not yet a robust quantitative method, available in the literature, which can reliably predict mortar strength based on its mix components. This limitation is due to the highly nonlinear relation between the mortar's compressive strength and the mixed components. In this paper, the application of artificial neural networks for predicting the compressive strength of mortars has been investigated. Specifically, surrogate models (such as artificial neural network models) have been used for the prediction of the compressive strength of mortars (based on experimental data available in the literature). Furthermore, compressive strength maps are presented for the first time, aiming to facilitate mortar mix design. The comparison of the derived results with the experimental findings demonstrates the ability of artificial neural networks to approximate the compressive strength of mortars in a reliable and robust manner.

• 제어로봇시스템학회:학술대회논문집
• /
• 제어로봇시스템학회 2003년도 ICCAS
• /
• pp.1036-1041
• /
• 2003

Non-destruction examinations are required in medical sciences and various engineering now. We wish to emulate the examinations in very simplified experiments. It is an educational program. We show a neural network analysis to predict the locations of a sound-source or a body irradiated by sound-waves in audio-region. The sound is an interest flux, and it enables to clear local-structures in a non-transparent space. However, the sound-propagation equations are not solved easily, therefore, we consider to adopt multi-layer neural-networks instead of the direct solutions. We used detected intensities and coordinates for input data and teaching data. A neural network learned them. The neural-network analysis decomposed the distance of 50cm. The resolution is rather rough; however, it is caused by the limitation of our equipments. Since there is no problem in the neural network processing, if we could revise experiments, then, progress of the resolution would be got. Thus, the proposed method functioned as an educational and simplified non-destruction examination.

• 제어로봇시스템학회논문지
• /
• 제21권4호
• /
• pp.378-384
• /
• 2015

Indoor localization is becoming one of the most important technologies for smart mobile applications with different requirements from conventional outdoor location estimation algorithms. Fingerprinting location estimation techniques based on neural networks have gained increasing attention from academia due to their good generalization properties. In this paper, we propose a novel location estimation algorithm based on an ensemble of multiple neural networks. The neural network ensemble has drawn much attention in various areas where one neural network fails to resolve and classify the given data due to its' inaccuracy, incompleteness, and ambiguity. To the best of our knowledge, this work is the first to enhance the location estimation accuracy in indoor wireless environments based on a neural network ensemble using fingerprinting training data. To evaluate the effectiveness of our proposed location estimation method, we conduct the numerical experiments using the TGn channel model that was developed by the 802.11n task group for evaluating high capacity WLAN technologies in indoor environments with multiple transmit and multiple receive antennas. The numerical results show that the proposed method based on the NNE technique outperforms the conventional methods and achieves very accurate estimation results even in environments with a low number of APs.

• 대한전기학회:학술대회논문집
• /
• 대한전기학회 1999년도 하계학술대회 논문집 B
• /
• pp.516-518
• /
• 1999

This paper presents an feed-forward neural network design instead PI controller for the speed control of an Induction Motor. The design employs the training strategy with Neural Network Controller(NNC) and Neural Network Emulator(NNE). Emulator identifies the motor by simulating the input and output map. In order to update the weights of the Controller. Emulator supplies the error path to the output stage of the controller using backpropagation algorithm. and then Controller produces an adequate output to the system due to neural networks learning capability. Therefore it becomes adjustable to the system with changing characteristics caused by a load. The speed control based on neural networks for induction motor is implemented by a vector controlled induction motor. The simulation results demonstrate that actual motor speed with neural network system well follows the reference speed minimizing the error and is available to implement on the vector control theory.

• 대한전기학회논문지:시스템및제어부문D
• /
• 제49권9호
• /
• pp.512-519
• /
• 2000

A Neural networks has been used for a expert system and fault diagnosis system. It is possible to nonlinear function mapping and parallel processing. Therefore It has been developing for a Diagnosis system of nuclear plower plant. In general Neural Networks is a static mapping but Dynamic Neural Network(DNN) is dynamic mapping.쪼두 a fault occur in system a state of system is changed with transient state. Because of a previous state signal is considered as a information DNN is better suited for diagnosis systems than static neural network. But a DNN has many weights so a real time implementation of diagnosis system is in need of a rapid network architecture. This paper presents a algorithm for RCP monitoring Alarm diagnosis system using Self Dynamic Neural Network(SDNN). SDNN has considerably fewer weights than a general DNN. Since there is no interlink among the hidden layer. The effectiveness of Alarm diagnosis system using the proposed algorithm is demonstrated by applying to RCP monitoring in Nuclear power plant.

• 한국유체기계학회 논문집
• /
• 제5권4호
• /
• pp.47-53
• /
• 2002

The steam generator feedwater flow-rate in a nuclear power plant was estimated by means of artificial neural networks with the wavelet analysis for enhanced information extraction. The fouling of venturi meters, used for steam generator feedwater flow-rate in pressurized water reactors, may result in unnecessary plant power derating. The back-propagation network was used to generate models of signals for a pressurized water reactor Multiple-input, single-output hetero-associative networks were used for evaluating the feedwater flow rate as a function of a set of related variables. The wavelet was used as a low pass filter eliminating the noise from the raw signals. The results have shown that possible fouling of venturi can be detected by neural networks, and the feedwater flow-rate can be predicted as an alternative to existing methods. The research has also indicated that the decomposition of signals by wavelet transform is a powerful approach to signal analysis for denoising.