• Journal of the Korean earth science society
• /
• v.30 no.3
• /
• pp.294-304
• /
• 2009

A statistical prediction model for the typhoon intensity and track in the Northwestern Pacific area was developed based on the artificial neural network scheme. Specifically, this model is focused on the 5-day prediction after tropical cyclone genesis, and used the CLIPPER parameters (genesis location, intensity, and date), dynamic parameters (vertical wind shear between 200 and 850hPa, upper-level divergence, and lower-level relative vorticity), and thermal parameters (upper-level equivalent potential temperature, ENSO, 200-hPa air temperature, mid-level relative humidity). Based on the characteristics of predictors, a total of seven artificial neural network models were developed. The best one was the case that combined the CLIPPER parameters and thermal parameters. This case showed higher predictability during the summer season than the winter season, and the forecast error also depended on the location: The intensity error rate increases when the genesis location moves to Southeastern area and the track error increases when it moves to Northwestern area. Comparing the predictability with the multiple linear regression model, the artificial neural network model showed better performance.

• Korean Chemical Engineering Research
• /
• v.52 no.6
• /
• pp.720-726
• /
• 2014

PEM fuel cell vehicles have been getting much attraction due to a sort of highly clean and effective transportation. The onboard fuel processor, however, is inevitably required to supply the hydrogen by conversion from some fuels since there are not enough available hydrogen stations nearby. A lot of studies have been focused on analyses of ATR reactor under the assumption of thermo-neutral condition and those of the optimized process for the minimization of energy consumption using thermal efficiency as an objective function, which doesn't guarantee the maximum hydrogen production. In this study, the analysis of optimization for 100 kW PEMFC onboard fuel processor was conducted targeting various fuels such as gasoline, LPG, diesel using newly defined hydrogen efficiency and keeping simply synthesized heat exchanger network regardless of external utilities leading to compactness and integration. Optimal result of gasoline case shows 9.43% reduction compared to previous study, which shows the newly defined objective function leads to better performance than thermal efficiency in terms of hydrogen production. The sensitivity analysis was also done for hydrogen efficiency, heat recovery of each heat exchanger, and the cost of each fuel. Finally, LPG was estimated as the most economical fuel in Korean market.

• Korean Journal of Remote Sensing
• /
• v.35 no.6_1
• /
• pp.999-1009
• /
• 2019

The importance of high-resolution sea ice maps of the Arctic Ocean is increasing due to the possibility of pioneering North Pole Routes and the necessity of precise climate prediction models. In this study,sea ice classification algorithms for two deep learning models were examined using Sentinel-1 A/B SAR data to generate high-resolution sea ice classification maps. Based on current ice charts, three classes (Open Water, First Year Ice, Multi Year Ice) of training data sets were generated by Arctic sea ice and remote sensing experts. Ten sea ice classification algorithms were generated by combing two deep learning models (i.e. Simple CNN and Resnet50) and five cases of input bands including incident angles and thermal noise corrected HV bands. For the ten algorithms, analyses were performed by comparing classification results with ground truth points. A confusion matrix and Cohen's kappa coefficient were produced for the case that showed best result. Furthermore, the classification result with the Maximum Likelihood Classifier that has been traditionally employed to classify sea ice. In conclusion, the Convolutional Neural Network case, which has two convolution layers and two max pooling layers, with HV and incident angle input bands shows classification accuracy of 96.66%, and Cohen's kappa coefficient of 0.9499. All deep learning cases shows better classification accuracy than the classification result of the Maximum Likelihood Classifier.

• KSII Transactions on Internet and Information Systems (TIIS)
• /
• v.14 no.11
• /
• pp.4290-4309
• /
• 2020

In recent years, advancements in machine learning capabilities have allowed it to see widespread adoption for tasks such as object detection, image classification, and anomaly detection. However, despite their promise, a limitation lies in the fact that a network's performance quality is based on the data which it receives. A well-trained network will still have poor performance if the subsequent data supplied to it contains artifacts, out of focus regions, or other visual distortions. Under normal circumstances, images of the same scene captured from differing points of focus, angles, or modalities must be separately analysed by the network, despite possibly containing overlapping information such as in the case of images of the same scene captured from different angles, or irrelevant information such as images captured from infrared sensors which can capture thermal information well but not topographical details. This factor can potentially add significantly to the computational time and resources required to utilize the network without providing any additional benefit. In this study, we plan to explore using image fusion techniques to assemble multiple images of the same scene into a single image that retains the most salient key features of the individual source images while discarding overlapping or irrelevant data that does not provide any benefit to the network. Utilizing this image fusion step before inputting a dataset into the network, the number of images would be significantly reduced with the potential to improve the classification performance accuracy by enhancing images while discarding irrelevant and overlapping regions.

• Nuclear Engineering and Technology
• /
• v.56 no.9
• /
• pp.3571-3584
• /
• 2024

Reactor core transient calculation is very important for the reactor safety analysis, in which the kernel is neutron kinetics calculation by simulating the variation of neutron density or thermal power over time. Compared with the point kinetics method, the time-space neutron kinetics calculation can provide accurate variation of neutron density in both space and time domain. But it consumes a lot of resources. It is necessary to develop a surrogate model that can quickly obtain the temporal and spatial variation information of neutron density or power with acceptable calculation accuracy. This paper uses the time-varying characteristics of power to construct a time function, parameterizes the time-varying characteristics which contains the information about the spatial change of power. Thereby, the amount of targets to predict in the space domain is compressed. A surrogate method using the machine learning is proposed in this paper. In the construction of a neural network, the input is processed by a convolutional layer, followed by a fully connected layer or a deconvolution layer. For the problem of time sequence disturbance, a structure combining convolutional neural network and recurrent neural network is used. It is verified in the tests of a series of 1D, 2D and 3D reactor models. The predicted values obtained using the constructed neural network models in these tests are in good agreement with the reference values, showing the powerful potential of the surrogate models.

• Journal of Adhesion and Interface
• /
• v.9 no.3
• /
• pp.7-13
• /
• 2008

The particulate type silicon carbide (SiC) and fiber type carbon fiber (CF) filler, of similar thermal conductivities, were mixed with polyetheretherketone (PEEK) to investigate the filler effects on the thermal diffusivity. The SiC and CF fillers had a good and uniform dispersion in PEEK matrix. Thermal diffusivities of PEEK composites were measured from ambient temperature up to $200^{\circ}C$ by laser flash method. The diffusivities were decreased as increasing temperature due to the phonon scattering between PEEK-filler and filler-filler interfaces. Thermal diffusivity of PEEK composites was increased with increasing filler content and the thermal conductivities of two-phase system were compared to the experimental results and it gave ideas on the filler dispersion, orientation, aspect ratio, and filler-filler interactions. Nielson equation gave a good prediction to the experimental results of PEEK/SiC. The easy network formation by CF was found to be substantially more effective than SiC and it gave a higher thermal diffusivities of PEEK/CF than PEEK/SiC.

• Journal of Surface Science and Engineering
• /
• v.50 no.6
• /
• pp.465-472
• /
• 2017

The ingot fabrication conditions related with the thermal shock bearing phase and microstructure have investigated for the rare earth zirconate ceramic material, lanthanum gadolinium zirconate, as a thermal barrier coating using electron beam evaporation method. The thermal shock resistance of the prepared ingot was evaluated by high energy electron beam irradiation. The rare earth zirconate ceramic powder was prepared by controlling the raw material powder composition of $La_2O_3$, $Gd_2O_3$ and $ZrO_2$ so as to have a composition of $(La_{0.3}Gd_{0.7})_2Zr_2O_7$ which was selected from the former study. Ingot samples were prepared under two conditions. The first condition is prepared by sintering the prepared powder mixture to form an ingot. The second condition is prepared by calcining the prepared powder mixture to form a composite phase and then sintering to form an ingot. X-ray diffraction(XRD) and Scanning Electron Microscope(SEM) were used to analyze phase forming behavior and microstructure of ingot samples. Nanoindentation method used to obtain elastic modulus and hardness of each ingot specimen. Also the stress distribution of ingot was simulated by using FEM method assuming the ingot surface was exposed to electron beam. As a results, in the case of an ingot having a network-shaped microstructure in which relatively coarse pores are included, it seems that the thermal shock resistance was higher than in the case of an ingot having a microstructure composed of relatively fine grains only or particles with the similar level size when the high energy electron beam irradiation.

• The Journal of the Acoustical Society of Korea
• /
• v.21 no.2E
• /
• pp.98-104
• /
• 2002

The present study describes a combination method to recognize the human affective states such as anger, happiness, sadness, or surprise. For this, we extracted emotional features from voice signals and facial expressions, and then trained them to recognize emotional states using hidden Markov model (HMM) and neural network (NN). For voices, we used prosodic parameters such as pitch signals, energy, and their derivatives, which were then trained by HMM for recognition. For facial expressions, on the other hands, we used feature parameters extracted from thermal and visible images, and these feature parameters were then trained by NN for recognition. The recognition rates for the combined parameters obtained from voice and facial expressions showed better performance than any of two isolated sets of parameters. The simulation results were also compared with human questionnaire results.

• Proceedings of the Membrane Society of Korea Conference
• /
• 1996.10a
• /
• pp.86-87
• /
• 1996

IPN (Interpenetrating Polymer Network) is a mixture of two or more crosslinked polymers with physically interlocked network structures between the component polymers. IPN can be classified as an alloy of thermosets and has the characteristics of thermosets such as the thermal resistance and chemical resistance and also has the characteristics of polymer alloys with enhanced impact resistance and amphoteric properties. The physical interlocking during the synthesis restricts the phase separation of the component polymer with chemical pinning process, thus the control of morphology is possible through variations of the reaction temperature and pressure, catalyst concentration and crosslinking agent concentration. Finely dispersed domain structure can be obtained through IPN synthesis of polymer components with gross immiscibility. In membrane applications, particularly for the separation of liquid mixtures, crosslinked polymer component with specific affinity to the permeate is needed. With the presence of the permeant-inert polymer component, the mechanical strength and the selectivity of the membranes are enhanced by restricting the swelling of the transporting polymer component networks.

• Proceedings of the KIEE Conference
• /
• 2005.10b
• /
• pp.138-140
• /
• 2005

A fuzzy ARTMAP neural network and a fuzzy ART neural network are proposed to identify $H_2S$, $NH_3$ and their mixtures and to estimate their concentrations, respectively. Features are extracted from a micro gas sensor array operated in a thermal modulation plan. After dimensions of the features are reduced by a preprocessing scheme, the features are fed into the proposed fuzzy neural networks. By computer simulations, the proposed methods are shown to be fast in learning and accurate in concentration estimating. The results are compared with other methods and discussed.