• Nuclear Engineering and Technology
• /
• 제51권3호
• /
• pp.702-708
• /
• 2019

This paper deals with break size estimation of loss of coolant accidents (LOCA) using a nonlinear autoregressive with exogenous inputs (NARX) neural network. Previous studies used static approaches, requiring time-integrated parameters and independent firing algorithms. NARX neural network is able to directly deal with time-dependent signals for dynamic estimation of break sizes in real-time. The case studied is a LOCA in the primary system of Bushehr nuclear power plant (NPP). In this study, number of hidden layers, neurons, feedbacks, inputs, and training duration of transients are selected by performing parametric studies to determine the network architecture with minimum error. The developed NARX neural network is trained by error back propagation algorithm with different break sizes, covering 5% -100% of main coolant pipeline area. This database of LOCA scenarios is developed using RELAP5 thermal-hydraulic code. The results are satisfactory and indicate feasibility of implementing NARX neural network for break size estimation in NPPs. It is able to find a general solution for break size estimation problem in real-time, using a limited number of training data sets. This study has been performed in the framework of a research project, aiming to develop an appropriate accident management support tool for Bushehr NPP.

• Smart Structures and Systems
• /
• 제26권2호
• /
• pp.175-184
• /
• 2020

Conventional Monte Carlo simulation-based methods for seismic risk assessment of water networks often require excessive computational time costs due to the hydraulic analysis. In this study, an Artificial Neural Network-based surrogate model was proposed to efficiently evaluate the flow-based system reliability of water distribution networks. The surrogate model was constructed with appropriate training parameters through trial-and-error procedures. Furthermore, a deep neural network with hidden layers and neurons was composed for the high-dimensional network. For network training, the input of the neural network was defined as the damage states of the k-dimensional network facilities, and the output was defined as the network system performance. To generate training data, random sampling was performed between earthquake magnitudes of 5.0 and 7.5, and hydraulic analyses were conducted to evaluate network performance. For a hydraulic simulation, EPANET-based MATLAB code was developed, and a pressure-driven analysis approach was adopted to represent an unsteady-state network. To demonstrate the constructed surrogate model, the actual water distribution network of A-city, South Korea, was adopted, and the network map was reconstructed from the geographic information system data. The surrogate model was able to predict network performance within a 3% relative error at trained epicenters in drastically reduced time. In addition, the accuracy of the surrogate model was estimated to within 3% relative error (5% for network performance lower than 0.2) at different epicenters to verify the robustness of the epicenter location. Therefore, it is concluded that ANN-based surrogate model can be utilized as an alternative model for efficient seismic risk assessment to within 5% of relative error.

• 제어로봇시스템학회:학술대회논문집
• /
• 제어로봇시스템학회 1992년도 한국자동제어학술회의논문집(국내학술편); KOEX, Seoul; 19-21 Oct. 1992
• /
• pp.243-247
• /
• 1992

This paper describes a neural network based control scheme with MRAC. The system consists of two neural network; one is for identifier and the other is for controller. Identification is firstly performed to learn the behavior of the nonlinear plant. Neural net controller is next trained by backpropagating the error at the output of plant through the identifier. Also the training method used in this paper repeatedly updates weights of neural network to track the reference model.

• 한국전산구조공학회논문집
• /
• 제16권3호
• /
• pp.229-236
• /
• 2003

손상평가를 위해 많은 연구자들에 의해 인공신경망이 이용되어 왔다. 그러나, 인공신경망을 이용한 손상평가에 있어 정확성과 능률성을 제고하기 위해서는 몇가지 문제점이 있다. 기존의 인공신경망 특히 역전파신경망(BPNN)의 경우 신경망 학습을 위해 많은 수의 학습패턴을 필요로 하며, 또한 신경망의 구조와 해의 수렴간에 어떤 확정적인 관계가 존재하지 않는다. 따라서 신경망의 은닉층의 수와 한 은닉층에서의 노드수는 시행착오적으로 결정되게 된다. 이러한 많은 훈련패턴의 준비와 최적의 신경망 구조 결정을 위해서는 많은 시간이 필요하다. 본 논문에서는 이러한 단점들을 극복하기 위해 확률신경망을 패턴분류기로 사용하였다. 이를 판형철도교의 손상평가에 수치해석적으로 검증하였다. 또한 확률신경망을 이용한 철도판형교 손상평가시 적절한 훈련패턴 선택을 위해 모드형상과 고유진동수를 사용한 경우의 적용성에 대해 검토하였다.

• 제어로봇시스템학회:학술대회논문집
• /
• 제어로봇시스템학회 1993년도 한국자동제어학술회의논문집(국내학술편); Seoul National University, Seoul; 20-22 Oct. 1993
• /
• pp.212-217
• /
• 1993

In this paper, the hardening depth in Laser surface hardening process is estimated using a multilayered neural network. Input data of the neural network are surface temperature of five points, power and travelling speed of Laser beam. A FDM(finite difference method) is used for modeling the Laser surface hardening process. This model is used to obtain the network's training data sample and to evaluate the performance of the neural network estimator. The simulational results showed that the proposed scheme can be used to estimate the hardening depth on real time.

• 대한전자공학회:학술대회논문집
• /
• 대한전자공학회 1999년도 하계종합학술대회 논문집
• /
• pp.500-503
• /
• 1999

In this paper, we analyze weight distributions of neural networks. If we construct a vector containing all weights of a neural network, then training process can be viewed as finding a solution point in the weight space. In order to obtain insight into the training process of neural networks, we investigate the distribution of the solution points in the weight space Experiments provide some interesting results, showing that solution points tend to form clusters in the weight space and the information may be used to speed up the training process.

• 한국조명전기설비학회:학술대회논문집
• /
• 한국조명전기설비학회 2006년도 춘계학술대회 논문집
• /
• pp.370-375
• /
• 2006

This paper suggests the method for estimating a partial discharge (PD) location using divide of the inside transformer as a grid. The PD location is found swiftly and economically compared with the typical method detecting a PD. The reason is that the location of PD is detected in the section. The estimation of PD location is trained using the Neural Network. JavaNNS(Java Neural Network Simulator) and SNNS(Stuttgart Neural Network Simulator) are used for searching the location of PD. The simulation procedure is following, The transformer is assumed that the case is a regular hexahedron. The sensor is installed in a proper location. A section of PD location is set as a target, and training set is studied with several PD locations in the inside of the transformer. As a result of training process, the learning capability of neural network is excellent. The PD location is detected by division of internal structure of transformer and application of neural network.

• 한국경영과학회:학술대회논문집
• /
• 한국경영과학회 1996년도 추계학술대회발표논문집; 고려대학교, 서울; 26 Oct. 1996
• /
• pp.27-30
• /
• 1996

Determination of input variables for artificial neural network (ANN) depends entirely on the judgement of a modeller. As the number of input variables increases, the training time for the resulting ANN increases exponentially. Moreover, larger number of input variables does not guarantee better performance. In this research, we employ Genetic Algorithm for selecting proper input variables that yield the best performance in training the resulting ANN.

• 한국지반공학회논문집
• /
• 제20권8호
• /
• pp.77-87
• /
• 2004

본 논문에서는 국내 서남해안 11개 지역에서 수행된 63회의 피에조콘 시험결과와 176개의 선행압밀하중 자료로부터 국내 연약지반의 선행압밀하중 예측을 위한 오차 역전파 알고리즘으로 학습된 피에조콘 인공신경망 모델을 구축하였다. 전체 자료 중 147개의 자료만이 인공신경망 모델 구축을 위한 학습과정에 사용되었으며 학습에 사용되지 않은 29개의 자료를 구축된 인공신경망의 검증에 활용하였다. 또한 기존의 경험모델 및 이론모델과 비교하여 제안된 인공신경망 모델의 유용성을 확인하였다. 연구를 통하여 4-4-9-1의 구조를 갖는 간단한 다층 인공신경망이 구축되었으며 입력값으로는 피에조콘 선단저항력 $q_T$, 관입간극수압 $u_2$그리고 지반의 총상재하중 $\sigma_{vo}$ 및 유효상재하중 $\sigma'_{vo}$ 이 사용되었다. 제안된 인공신경망 모델은 학습되지 않은 새로운 검증자료에 대한 예측을 통하여 입력변수들과 선행압밀 하중 간의 비선형적 상관관계를 성공적으로 모델하는 것으로 검증되었으며 정확성면에서는 기존의 이론모델과 국내외 경험모델과 비교할 때 월등히 향상된 예측능력을 가진 것으로 나타났다. 뿐만 아니라 제안된 모델은 국내 특정지 역에 대한 모델이 아니라 서남해안의 다양한 지반특성을 갖는 지반에서 수행된 자료를 바탕으로 구축되어 데이터베이스에 포함되지 않은 지역에 대하여도 매우 타당성있는 예측결과를 주어 특정지역에 국한된 지역의존적 예측이 아닌 일반화된 지역에서 적용할 수 있을 것으로 판단된다.

• 한국소음진동공학회논문집
• /
• 제17권7호
• /
• pp.662-667
• /
• 2007

A new neuro-control scheme for active control of structures is proposed. It utilizes lattice pattern of state vector as training data of probabilistic neural network(PNN). Therefore. it is the so-called lattice probabilistic neural network(LPNN). PNN makes control forces by using all the training patterns. Therefore, it takes much time to obtain a control force in application. This inevitably may delay the control action. However. control force of LPNN is calculated by using only the adjacent information of LPNN input. So, the response of LPNN is greatly faster than PNN. The proposed control algorithm is applied for three story building under California and El Centro earthquakes. Also, control results of the LPNN are compared with those of the conventional PNN. The structural responses have been suppressed effectively by the proposed algorithm.