• 한국정보과학회논문지:소프트웨어및응용
• /
• 제34권7호
• /
• pp.585-594
• /
• 2007

오류역전파 방법을 이용하는 신경망들은 패턴들의 학습시간이 매우 오래 걸리고 또한 추가학습과 반복학습의 한계를 가지며, 이런 단점을 보완할 수 있는 이진신경망(Binary Neural Network, BNN)이 Aleksander에 의해 제안되었다. 그러나 BNN도 반복학습에 있어서는 단점을 가지고 있으며, 일반화 패턴을 추출하기 어렵다. 본 논문에서는 BNN의 구조를 개선하여 반복학습과 추가학습이 가능할 뿐 아니라, 특징점들까지 추출할 수 있는 다중 판별자를 가지는 삼차원 뉴로 시스템을 제안한다. 제안된 모델은 기존의 BNN을 기반으로 하여 만들어진 이차원 특징을 가지는 Single Layer Network(SLN)에 귀환회로가 추가되어 특징점들을 누적할 수 있는 삼차원 신경망이다. 학습을 통해 누적된 정보는 판별자의 각 신경세포에 임계치를 조정함으로써 일반화 패턴을 추출할 수 있다. 그리고 생성된 일반화 패턴을 인식에 재사용함으로써 반복학습의 효율성을 높였다. 최종 판정 단계에서는 Maximum Response Detector(MRD)를 이용하였다. 본 논문에서 제안한 시스템을 평가하기 위하여 NIST에서 제공하는 숫자 자료를 이용하였으며, 99.3%의 인식률을 얻었다.

• Nuclear Engineering and Technology
• /
• 제48권2호
• /
• pp.482-494
• /
• 2016

The Feshbache-Kermane-Koonin (FKK) multi-step direct (MSD) theory of pre-equilibrium reactions has been used to compute the single-step cross-sections for some (p,${\alpha}$) reactions using the knock-on and pick-up reaction mechanisms at two incident proton energies. For the knock-on mechanism, the reaction was assumed to have taken place by the direct ejection of a preformed alpha cluster in a shell-model state of the target. But the reaction was assumed to have taken place by the pick-up of a preformed triton cluster (also bound in a shell-model state of the target core) by the incident proton for the pick-up mechanism. The Yukawa forms of potential were used for the proton-alpha (for the knock-on process) and proton-triton (for the pick-up process) interaction and several parameter sets for the proton and alpha-particle optical potentials. The calculated cross-sections for both mechanisms gave satisfactory fits to the experimental data. Furthermore, it has been shown that some combinations of the calculated distorted wave Born approximation cross-sections for the two reaction mechanisms in the FKK MSD theory are able to give better fits to the experimental data, especially in terms of range of agreement. In addition, the theory has been observed to be valid over a wider range of energy.

• Nuclear Engineering and Technology
• /
• 제53권8호
• /
• pp.2696-2707
• /
• 2021

The purpose of this study is to investigate the efficiency of various structural modeling schemes for evaluating seismic performances and fragility of the reactor containment building (RCB) structure in the advanced power reactor 1400 (APR1400) nuclear power plant (NPP). Four structural modeling schemes, i.e. lumped-mass stick model (LMSM), solid-based finite element model (Solid FEM), multi-layer shell model (MLSM), and beam-truss model (BTM), are developed to simulate the seismic behaviors of the containment structure. A full three-dimensional finite element model (full 3D FEM) is additionally constructed to verify the previous numerical models. A set of input ground motions with response spectra matching to the US NRC 1.60 design spectrum is generated to perform linear and nonlinear time-history analyses. Floor response spectra (FRS) and floor displacements are obtained at the different elevations of the structure since they are critical outputs for evaluating the seismic vulnerability of RCB and secondary components. The results show that the difference in seismic responses between linear and nonlinear analyses gets larger as an earthquake intensity increases. It is observed that the linear analysis underestimates floor displacements while it overestimates floor accelerations. Moreover, a systematic assessment of the capability and efficiency of each structural model is presented thoroughly. MLSM can be an alternative approach to a full 3D FEM, which is complicated in modeling and extremely time-consuming in dynamic analyses. Specifically, BTM is recommended as the optimal model for evaluating the nonlinear seismic performance of NPP structures. Thereafter, linear and nonlinear BTM are employed in a series of time-history analyses to develop fragility curves of RCB for different damage states. It is shown that the linear analysis underestimates the probability of damage of RCB at a given earthquake intensity when compared to the nonlinear analysis. The nonlinear analysis approach is highly suggested for assessing the vulnerability of NPP structures.

• Nuclear Engineering and Technology
• /
• 제53권12호
• /
• pp.4098-4105
• /
• 2021

The aim of this study is to calculate the JO-IMRT dose distributions based on the AAPM TG-119 using Monte Carlo (MC) simulation and Prowess Panther treatment planning system (TPS) (Panther, Prowess Inc., Chico, CA). JO-IMRT dose distributions of AAPM TG-119 were calculated by the TPS and were recalculated by MC simulation. The DVHs and 3D gamma index using global methods implemented in the PTW-VeriSoft with 3%/3 mm were used for evaluation. JO-IMRT dose distributions calculated by TPS and MC were matched the TG-119 goals. The gamma index passing rates with 3%/3 mm were 98.7% for multi-target, 96.0% for mock prostate, 95.4% for mock head-and-neck, and 96.6% for C-shape. The dose in the planning target volumes (PTV) for TPS was larger than that for the MC. The relative dose differences in D99 between TPS and MC for multi-target are 1.52%, 0.17% and 1.40%, for the center, superior and inferior, respectively. The differences in D95 are 0.16% for C-shape; and 0.06% for mock prostate. Mock head-and-neck difference is 0.40% in D99. In contrast, the organ curve for TPS tended to be smaller than MC values. JO-IMRT dose distributions for the AAPM TG-119 calculated by the TPS agreed well with the MC.

• 한국건축시공학회지
• /
• 제18권6호
• /
• pp.621-631
• /
• 2018

국내 공공임대주택에서 유지관리 단계에서의 수선유지비는 효율적 관리의 필요와 함께 그 중요성이 부각되고 있다. 본 연구에서는 수선유지비를 예측하는 방법으로서 기존의 실적자료를 통한 방법을 대신하여 물량기반의 예측방법을 제안하고자 한다. 견적방식 모델을 통하여 공동주택 유지관리단계의 40년간을 대상으로 연차 단위별로 수선유지비용을 예측하고 분포 특성에 대한 정보를 제공한다. 수선유지비예측의 정확성을 제고하기 위해, 최근 변경된 장기수선항목 및 수선주기를 반영하였다. 또한 최근에 건설된 공동주택 마감수준을 수선대상으로 포함하였다. 수량산출은 공공임대주택 5개 사례현장을 선정하여 수행되었으며, 세대수 및 연면적 단위로 환산하여 공공임대주택의 수선유지비를 범용적으로 예측하는데 활용될 수 있도록 분석하였다.

• Nuclear Engineering and Technology
• /
• 제53권5호
• /
• pp.1446-1453
• /
• 2021

In the past, several experimental investigations aiming at characterizing the natural circulation (NC) behavior in test facilities were carried out. They showed a variety of flow patterns characterized by an inverted U-shape of the NC flow curve versus primary mass inventory. On the other hand, attempts to reproduce such curves using thermal-hydraulic system codes, showed 10-30% differences between the measured and calculated NC mass flow rate. Actually, the used computer codes are generally based upon nodalization using single U-tube representation. Such model may not allow getting accurate simulation of most of the NC phenomena occurring during such tests (like flow redistribution and flow reversal in some SG U-tubes). Simulations based on multi-U-tubes model, showed better agreement with the overall behavior, but remain unable to predict NC phenomena taking place in the steam generator (SG) during the experiment. In the current study, the CATHARE code is considered in order to assess a NC characterization test performed in the four loops PKL facility. For this purpose, four different SG nodalizations including, single and multi-U-tubes, 1D and 3D SG inlet/outlet zones are considered. In general, it is shown that the 1D and 3D models exhibit similar prediction results up to a certain point of the rising part of the inverted U-shape of the NC flow curve. After that, the results bifurcate with, on the one hand, a tendency of the 1D models to over-predict the measured NC mass flow rate and on the other hand, a tendency of the 3D models to under-predict the NC flow rate.

• 한국컴퓨터정보학회논문지
• /
• 제26권1호
• /
• pp.57-67
• /
• 2021

최근 딥 러닝(Deep Learning) 분석에 이질적인 데이터를 함께 사용하는 멀티모달(Multi-modal) 딥러닝 기술이 많이 활용되고 있으며, 특히 텍스트로부터 자동으로 이미지를 생성해내는 Text to Image 합성에 관한 연구가 활발하게 수행되고 있다. 이미지 합성을 위한 딥러닝 학습은 방대한 양의 이미지와 이미지를 설명하는 텍스트의 쌍으로 구성된 데이터를 필요로 하므로, 소량의 데이터로부터 다량의 데이터를 생성하기 위한 데이터 증강 기법이 고안되어 왔다. 텍스트 데이터 증강의 경우 유의어 대체에 기반을 둔 기법들이 다수 사용되고 있지만, 이들 기법은 명사 단어의 유의어 대체 시 이미지의 내용과 상이한 텍스트를 생성할 가능성이 있다는 한계를 갖는다. 따라서 본 연구에서는 단어가 갖는 품사별 특징을 활용하는 텍스트 데이터 증강 방안, 즉 일부 품사에 대해 단어 계층 정보를 활용하여 단어를 대체하는 방안을 제시하였다. 또한 제안 방법론의 성능을 평가하기 위해 MSCOCO 데이터를 사용하여 실험을 수행하여 결과를 제시하였다.

• Nuclear Engineering and Technology
• /
• 제54권6호
• /
• pp.2023-2030
• /
• 2022

A novel pulse fitting analysis (PFA) method is presented for the acquisition of nuclear spectra. The charging process of the feedback capacitor in the resistive feedback charge-sensitive preamplifier is equivalent to the impulsive pulse, and its impulse response function (IRF) can be obtained by non-linear fitting of the falling edge of the nuclear pulse. The integral of the IRF excluding the baseline represents the energy deposition of the particles in the detector. In addition, since the non-linear fitting process in PFA method is difficult to achieve in the conventional architecture of spectroscopy system, a new multi-channel analyzer (MCA) based on Zynq SoC is proposed, which transmits all the data of nuclear pulses from the programmable logic (PL) to the processing system (PS) by high-speed AXI-Stream in order to implement PFA method with precision. The linearity of new MCA has been tested. The spectrum of ¹³⁷Cs was obtained using LaBr₃(Ce) scintillator detector, and was compared with commercial MCA by ORTEC. The results of tests indicate that the MCA based on PFA method has the same performance as the commercial MCA based on pulse height analysis (PHA) method and excellent linearity for γ-rays with different energies, which infers that PFA method is an effective and promising method for the acquisition of spectra. Furthermore, it provides a new solution for nuclear pulse processing algorithms involving regression and iterative processes.

• Nuclear Engineering and Technology
• /
• 제54권10호
• /
• pp.3717-3723
• /
• 2022

For the first time, two-dimensional temporal behavior of the edge localized mode (ELM) filament is measured in the edge tokamak plasma with a multi-channel electrical probe array (MCEP). MCEP, which has 16 floating probes (4 × 4), is mounted at the far scrape-off layer (SOL) region in the KSTAR. An electron temperature and an ion flux are measured by sideband method (SBM), which can achieve two-dimensional measurements with high time resolution. Furthermore, temporal evolutions of the electron temperature and the ion flux are obtained during the ELM occurrence. In the H-mode period, short spikes from ELM bursts are observed in measured plasma parameters, and the trend is similar to that of typical Hα signal. Interestingly, when blob-like ELM filaments crash the probe, the heat flux is significantly higher in a local region of the probe array. The results show that our probe array using the SBM can measure the ELM behavior and the plasma parameters without the effect of the stray current caused by the huge device. This study can provide valuable data needed to understand the interaction between the SOL plasma and the plasma facing components (PFCs).

• Nuclear Engineering and Technology
• /
• 제55권9호
• /
• pp.3150-3163
• /
• 2023

To improve the performance of blanket: maximizing the tritium breeding rate (TBR) for tritium self-sufficiency, and minimizing the Dose of backplate for radiation protection, most previous studies are based on manual corrections to adjust the blanket structure to achieve optimization design, but it is difficult to find an optimal structure and tends to be trapped by local optimizations as it involves multiphysics field design, which is also inefficient and time-consuming process. The artificial intelligence (AI) maybe is a potential method for the optimization design of the blanket. So, this paper aims to develop an intelligent optimization method based on an improved multi-objective NSGA-III algorithm and an adaptive BP neural network to solve these problems mentioned above. This method has been applied on optimizing the radial arrangement of a conceptual design of CFETR HCSB blanket. Finally, a series of optimal radial arrangements are obtained under the constraints that the temperature of each component of the blanket does not exceed the limit and the radial length remains unchanged, the efficiency of the blanket optimization design is significantly improved. This study will provide a clue and inspiration for the application of artificial intelligence technology in the optimization design of blanket.