• Bulletin of the Korean Chemical Society
• /
• v.28 no.9
• /
• pp.1472-1476
• /
• 2007

Artificial neural networks (ANNs) are successfully developed for the modeling and prediction of normalized polarity parameter (ETN) of 216 various solvents with diverse chemical structures using a quantitative-structure property relationship. ANN with architecture 5-9-1 is generated using five molecular descriptors appearing in the multi-parameter linear regression (MLR) model. The most positive charge of a hydrogen atom (q+), total charge in molecule (qt), molecular volume of solvent (Vm), dipole moment (μ) and polarizability term (πI) are input descriptors and its output is ETN. It is found that properly selected and trained neural network with 192 solvents could fairly represent the dependence of normalized polarity parameter on molecular descriptors. For evaluation of the predictive power of the generated ANN, an optimized network is applied for prediction of the ETN values of 24 solvents in the prediction set, which are not used in the optimization procedure. Correlation coefficient (R) and root mean square error (RMSE) of 0.903 and 0.0887 for prediction set by MLR model should be compared with the values of 0.985 and 0.0375 by ANN model. These improvements are due to the fact that the ETN of solvents shows non-linear correlations with the molecular descriptors.

• Nuclear Engineering and Technology
• /
• v.55 no.11
• /
• pp.4048-4056
• /
• 2023

The feasibility of using the Monte Carlo code MCS to generate multigroup cross sections for nodal diffusion simulations RAST-F of liquid metal fast reactors is investigated in this paper. The performance of the MCS/RAST-F code system is assessed using steady-state simulations of the ANTS-100e core. The results show good agreement between MCS/RAST-F and MCS reference solutions, with a keff difference of less than 77 pcm and root-mean-square differences in radial and axial power of less than 0.5% and 0.25%, respectively. Furthermore, the MCS/RAST-F reactivity feedback coefficients are within three standard deviations of the MCS coefficients. To validate the internal thermal-hydraulic (TH) feedback capability in RAST-F code, the coupled neutronic/TH1D simulation of ANTS-100e is performed using the case matrix obtained from MCS branch calculations. The results are compared to those obtained using the MARS-LBE system code and show good agreement with relative temperature differences in fuel and coolant of less than 0.8%. This study demonstrates that the MCS/RAST-F code system can produce accurate results for core steady-state neutronic calculations and for coupled neutronic/TH simulations.

• KSII Transactions on Internet and Information Systems (TIIS)
• /
• v.16 no.11
• /
• pp.3780-3797
• /
• 2022

The estimation of radio map (RM) is a fundamental and critical task for the network planning and optimization performance of mobile communication. In this paper, a RM estimation method is proposed based on a deep dual learning structure. This method can simultaneously and accurately reconstruct the urban building map (UBM) and estimate the RM of the whole cell by only part of the measured reference signal receiving power (RSRP). Our proposed method implements UBM reconstruction task and RM estimation task by constructing a dual U-Net-based structure, which is named RadioCycle. RadioCycle jointly trains two symmetric generators of the dual structure. Further, to solve the problem of interference negative transfer in generators trained jointly for two different tasks, RadioCycle introduces a dynamic weighted averaging method to dynamically balance the learning rate of these two generators in the joint training. Eventually, the experiments demonstrate that on the UBM reconstruction task, RadioCycle achieves an F1 score of 0.950, and on the RM estimation task, RadioCycle achieves a root mean square error of 0.069. Therefore, RadioCycle can estimate both the RM and the UBM in a cell with measured RSRP for only 20% of the whole cell.

• Korean Journal of Air-Conditioning and Refrigeration Engineering
• /
• v.27 no.11
• /
• pp.596-602
• /
• 2015

In this study, experiments using bioelectronic signals and questionnaire surveys were carried out in learning conditions when temperatures changed from low- and high-uncomfortable to comfortable. As a result, the stress factor Photoplethysmography (PPG) decreased, while the Root Mean Square of Standard Deviation (RMSSD) of PPG increased when the indoor temperature was changed from low- or high-uncomfortable to comfortable. Additionally, the absolute power of the ${\alpha}$-wave in the brain increased. According to the analysis of the association between the questionnaire and bioelectronic signals, the standard deviation of the stress factor as measured by pulse was closely related to the result of the thermal sensation questionnaire. In addition, it was found that the concentration on studying improved under comfortable temperatures when compared to uncomfortable temperatures.

• Nuclear Engineering and Technology
• /
• v.53 no.6
• /
• pp.1796-1809
• /
• 2021

The subchannel of a research reactor used to generate high power density is designed to be narrow and rectangular and comprises plate-type fuels operating under downward flow conditions. Critical heat flux (CHF) is a crucial parameter for estimating the safety of a nuclear fuel; hence, this parameter should be accurately predicted. Here, machine learning is applied for the prediction of CHF in a narrow rectangular channel. Although machine learning can effectively analyze large amounts of complex data, its application to CHF, particularly for narrow rectangular channels, remains challenging because of the limited flow conditions available in existing experimental databases. To resolve this problem, we used four CHF correlations to generate pseudo-data for training an artificial neural network. We also propose a network architecture that includes pre-training and prediction stages to predict and analyze the CHF. The trained neural network predicted the CHF with an average error of 3.65% and a root-mean-square error of 17.17% for the test pseudo-data; the respective errors of 0.9% and 26.4% for the experimental data were not considered during training. Finally, machine learning was applied to quantitatively investigate the parametric effect on the CHF in narrow rectangular channels under downward flow conditions.

• Journal of the Korean Magnetic Resonance Society
• /
• v.17 no.1
• /
• pp.11-18
• /
• 2013

Rapid advances of computational power and method have made it practical to apply the time-consuming calculations with all-atom force fields and sophisticated potential energies into refining NMR structure. Added to the all-atom force field, generalized-Born implicit solvent model (GBIS) contributes substantially to improving the qualities of the resulting NMR structures. GBIS approximates the effects that explicit solvents bring about even with fairly reduced computational times. Although GBIS is employed in the final stage of NMR structure calculation with experimental restraints, the effects by GBIS on structures have been reported notable. However, the detailed effect is little studied in a quantitative way. In this study, we report GBIS refinements of ubiquitin and GB1 structures by six GBIS models of AMBER package with experimental distance and backbone torsion angle restraints. Of GBIS models tested, the calculations with igb=7 option generated the closest structures to those determined by X-ray both in ubiquitin and GB1 from the viewpoints of root-mean-square deviations. Those with igb=5 yielded the second best results. Our data suggest that the degrees of improvements vary under different GBIS models and the proper selection of GBIS model can lead to better results.

• Journal of Sensor Science and Technology
• /
• v.21 no.5
• /
• pp.352-358
• /
• 2012

A sensor network system can be an efficient tool for healthcare telemetry for multiple users due to its power efficiency. One drawback is its limited data size. This paper proposed a real-time application of data compression/decompression method in u-Health monitoring system in order to improve the network efficiency. Our high priority was given to maintain a high quality of signal reconstruction since it is important to receive undistorted waveform. Our method consisted of down sampling coding and differential Huffman coding. Down sampling was applied based on the Nyquist-Shannon sampling theorem and signal amplitude was taken into account to increase compression rate in the differential Huffman coding. Our method was successfully tested in a ZigBee and WLAN dual network. Electrocardiogram (ECG) had an average compression ratio of 3.99 : 1 with 0.24% percentage root mean square difference (PRD). Photoplethysmogram (PPG) showed an average CR of 37.99 : 1 with 0.16% PRD. Our method produced an outstanding PRD compared to other previous reports.

• Nuclear Engineering and Technology
• /
• v.55 no.2
• /
• pp.756-768
• /
• 2023

Neutron transport calculations are extremely challenging due to the high computational cost of large and complex problems. A multilevel octree grid algorithm (MLTG) of discrete ordinates method was developed to improve the modeling accuracy and simulation efficiency on 3-D Cartesian grids. The Balakovo-3 VVER-1000 neutron dosimetry benchmark is calculated to verify and validate this numerical technique. A simplified S₂ synthetic acceleration is used in the MLTG calculation method to improve the convergence of the source iterations. For the triangularly arranged fuel pins, we adopt a source projection algorithm to generate pin-by-pin source distributions of hexagonal assemblies. MLTG provides accurate geometric modeling and flexible fixed source description at a lower cost than traditional Cartesian grids. The total number of meshes is reduced to 1.9 million from the initial 9.5 million for the Balakovo-3 model. The numerical comparisons show that the MLTG results are in satisfactory agreement with the conventional S_N method and experimental data, within the root-mean-square errors of about 4% and 10%, respectively. Compared to uniform fine meshing, approximately 70% of the computational cost can be saved using the MLTG algorithm for the Balakovo-3 computational model.

• Wind and Structures
• /
• v.17 no.6
• /
• pp.609-627
• /
• 2013

The Yingxian wooden tower in China is currently the tallest wooden tower in the world. It was built in 1056 AD and is 65.86 m high. Field measurements of wind speed and wind-induced response of this tower are conducted. The wind characteristics, including the average wind speed, wind direction, turbulence intensity, gust factor, turbulence integral length scale and velocity spectrum are investigated. The power spectral density and the root-mean-square wind-induced acceleration are analyzed. The structural modal parameters of this tower are identified with two different methods, including the Empirical Mode Decomposition (EMD) combined with the Random Decrement Technique (RDT) and Hilbert transform technique, and the stochastic subspace identification (SSI) method. Results show that strong wind is coming predominantly from the West-South of the tower which is in the same direction as the inclination of the structure. The Von Karman spectrum can describe the spectrum of wind speed well. Wind-induced torsional vibration obviously occurs in this tower. The natural frequencies identified by EMD, RDT and Hilbert Transform are close to those identified by SSI method, but there is obvious difference between the identified damping ratios for the first two modes.

• Proceedings of the KIEE Conference
• /
• 2005.10c
• /
• pp.17-19
• /
• 2005

Variety of methods were discussed to reduce the cogging force in tubular permanent magnet type linear single phase AC generator. In paticular, the proposed methods depend on the variations of the permanent magnet construction. These methods Include two approaches in the form of sloped magnets, and conical magnets in addition to the conventional method of varying the magnet length. The undesired cogging force ripples were calculated by a two dimensional Finite Element Method(FEM). Moreover, the generated electromotive force in the stator coils was calculated fur each configuration of the permanent magnet. The experimental results agreed well with those obtained from the FEM-based simulations. Sufficient reduction in the cogging force was achieved over the range of 40% while the root mean square of the output voltage was maintained. It was found that the sloping the permanent magnet decreased the cogging force and at the same time increased the generated rms voltage of the AC generator. The performance of the designed linear AC generator was evaluated in terms of its efficiency, total weight, losses, and power to weight ratio.