• The Transactions of the Korea Information Processing Society
• /
• v.13 no.9
• /
• pp.444-452
• /
• 2024

This study proposes a deep learning architecture optimized for fire detection derived through Layer Importance Evaluation. In order to solve the problem of unnecessary complexity and operation of the existing Convolutional Neural Network (CNN)-based fire detection system, the operation of the inner layer of the model based on the weight and activation values was analyzed through the Layer Importance Evaluation technique, the layer with a high contribution to fire detection was identified, and the model was reconstructed only with the identified layer, and the performance indicators were compared and analyzed with the existing model. After learning the fire data using four transfer learning models: Xception, VGG19, ResNet, and EfficientNetB5, the Layer Importance Evaluation technique was applied to analyze the weight and activation value of each layer, and then a new model was constructed by selecting the top rank layers with the highest contribution. As a result of the study, it was confirmed that the implemented architecture maintains the same performance with parameters that are about 80% lighter than the existing model, and can contribute to increasing the efficiency of fire monitoring equipment by outputting the same performance in accuracy, loss, and confusion matrix indicators compared to conventional complex transfer learning models while having a learning speed of about 3 to 5 times faster.

• Proceedings of the Korean Institute of Navigation and Port Research Conference
• /
• 2004.11a
• /
• pp.81-87
• /
• 2004

It includes the considerable concentration and dangerousness as the peculiar work of a pilot and dock-master to berthing/deberthing the big object as for the ship within the port or the ship yard. A tug utilization is getting increased in this berthing/deberthing work and the own ship is affected a lot by external force due to moving with low advance speed. In this study, we constructed the 2 dimension virtual system which can conduct the berthing/deberthing manoeuvring work by using mainly tugs in a external force, particularly strong wind. Also, propose objective standard that could estimate the degree of manoeuvring difficulty, and conducted simulation experiment for this. we analyzed correlations between the subjective estimation which is described numerically the decreased dangerousness and the objective index which is related to the main parameter regarding manoeuvring by using this simulator from the result of conducting simulation experiment. And then we discussed the estimation technique of manoeuvring difficulty.

• Journal of the Korean Chemical Society
• /
• v.37 no.5
• /
• pp.503-512
• /
• 1993

The purpose of this study was to investigate the retention behavior of phenols and to predict their retention in RPLC. The retention data of twenty-five phenols were measured on a $\mu-{Bondapak}\;C_{18}$ and a polymeric $C_{18}$ columns with methanol-water and acetonitrile-water as a mobile phase. From the observation of enthalpy-entropy compensation phenomenon, the following conclusions are drawn with regard to the retention mechanism: 1) the retention mechanism of nitrophenols in different from that of metheyl-and chlorophenols in both mobile phase; 2) in methanol-water mobile phase, the retention mechanism of methyl-and chlorophenols is consistent in the range of methanol-water composition; 3) on the other hand, in the case of acetonitrile-water mobile phase, the retention mechanism depends on the volume fraction of acetonitrile. It means that the retention mechanism can not be explained only by a simple interaction. Based on retention data as compared with two columns, it may be said that the hydrophobic interaction of phenols with polymeric $C_{18}$ column was greater than that with monomeric $C_{18}$ column. The equations for predicting the retention of phenols were derived by using hydrophobic substituent constant $(\pi)$ and the sum of Hammett's constant $(\sigma)$ and Taft's steric constant $(E_s)$.

• Journal of Korean Tunnelling and Underground Space Association
• /
• v.10 no.4
• /
• pp.397-404
• /
• 2008

Excessive amount of groundwater flowed into tunnel, while constructing Incheon international airport railway. Tunnel passes under subway line no. 2 with only 1.76 m below. To protect the existing structure, TRcM excavation method was applied. As station and construction shaft are already constructed, which are located back and forth of TRcM section, 86.4 ton per day of groundwater inflow is against expectation. To identify mechanism of excessive water inflow, hydraulic back analyses were performed. Then, hydro-mechanical coupled analysis were also performed with the hydrogeologic parameters identified, whose results are investigated for checking the stability of adjacent structures to the tunnel under construction. And a number of mechanical analyses were also performed to check the hydro-mechanical coupling effect. The result from the mechanical analysis shows that subsidence and tunnel ceiling displacement will be 0.85 mm and 1.32 mm. The result of hydro-mechanical couple analysis shows that subsidence and maximum tunnel ceiling displacement will be 1.2 mm and 1.72 mm. Additional displacements caused by groundwater draw down were identified, however, displacement is minute.

• KSCE Journal of Civil and Environmental Engineering Research
• /
• v.32 no.4D
• /
• pp.407-419
• /
• 2012

It is difficult to obtain the accurate and homogeneous height information over the whole Korea due to the effect of different vertical datums have been divided into land and sea part. In this study, we tried to unify the different vertical datums using the precise spirit leveling between TBM (tidal bench mark) and BM (bench mark) in order to solve the problems caused by different vertical datums. For this, the vertical datum offsets at observed points which were calculate from leveling results and then transformation model of vertical datum will be modelled using calculated offsets along the coastal line. For suggesting the precise modelling method to vertical datum transformation, we analyzed results from various interpolation methods such as Spline and LSC method. As the results from analysis, the LSC method combined with 4-parameters trend model is more suitable for modelling the offsets between vertical datums. The final transformation model of vertical datum using the combination of LSC and 4-parameter model which provides the transformation accuracies of ${\pm}10.4cm{\sim}14.8cm$ level. And, the software for vertical datum transformation that was also developed using the final model in order to convert the height information included in various spatial data effectively. Therefore, the transformation model between vertical datums of land and sea part, which is developed in this study, is expected to minimize the confusion caused by mismatch of height information in the use of spatial data, and it also can be minimize economic and time losses in various application fields such as coastal development project, coastal disaster prevention, etc.

• Journal of the Korean Association of Geographic Information Studies
• /
• v.9 no.1
• /
• pp.1-12
• /
• 2006

Approaches to spatial analysis differ from the type of spatial objects to be treated. Especially, in here, the case where two spatial data sets coexist is considered. The goal of such case lies on detecting a subset of spatial objects out of a large set that affects the distribution of the other object. However, it is not easy to extract a subset from a large set by visualization just with the help of GIS since huge amount of data are provided nowadays. In this research, therefore, relationship between two different spatial data are analyzed by quantitative measure in the case study of marketing geography. A purchase history data of a small retail store and the location of its competitors are given as source data for the analysis. The goal of analysis from the aspect of this case study is to extract strong competitors of the store that affects the sales amount of the store among many competitors. With the result, therefore, it is expected that market area pattern and competitive power of stores under micro scale retail environment would be understood by quantitative measure.

• Journal of Digital Convergence
• /
• v.14 no.5
• /
• pp.333-339
• /
• 2016

Technology to optimize and utilize the use and supply of the electric power between consumer and supplier has been on the rise among the smart grid power market network in electric power demand management based on the Internet. Open Automated Demand Response system protocol, which can deliver Demand Response needed in electric power demand management to electricity supplier, system supplier and even the user is openADR 2.0b. This paper used the most credible, cosmopolitanly proliferated EPRI open source and analysed the variety of security vulnerability that developed VEN and VTN system may have. Using the simulator for attacking openADR protocol, the VEN/VTN system that has been implemented as EPRI open source was conducted to attack in a variety of ways. As a result of the analysis, we were able to get the results that the VEN/VTN system has security vulnerabilities to the parameter tampering attacks and service flow falsification attack. In conclusion, if you want to implement the openADR2.0b protocol system in the open or two-way communication environment smart grid network, considering a variety of security vulnerability should be sure to seek security technology and services.

• Journal of the Korean Chemical Society
• /
• v.26 no.6
• /
• pp.363-368
• /
• 1982

Ultraviolet spectrophotometric investigations were carried out on the systems of pyridine with iodine, iodine monobromide and iodine monochloride in carbon tetrachloride. The results reveal the formation of the one to one molecular complexes of the type, $C_5H_5N{\cdot}I_2$, $ C_5H_5$N{\cdot}IBr and $ C_5H_5N{\cdot}ICl$. Considering ${\lambda}_max$ according to the formation of charge transfer complexes has the blue shift with the increasing temperatures$25, 40, 60^{\circ}C$ the equilibrium constants K and molar absorptivities $\varepsilon$ of complexes were obtained. From these values, the thermodynamic parameters ${\Delta}H$, ${\Delta}G$ and ${\Delta}S$ for the formation of the above charge transfer complexes were obtained. These results indicate that the relative stabilities of iodine, iodine monobromide and iodine monochloride complexes with pyridine increase in the order, $ C_5H_5N{\cdot}I_2$ < $ C_5H_5N{\cdot}IBr$ <$ C_5H_5N{\cdot}ICl$. This may be a measure of relative acidity of halogen and interhalogen toward pyridine and can be explained by the polarizabilities of electron acceptors and the difference of electronegativities of halogen atoms.

• Journal of IKEEE
• /
• v.22 no.1
• /
• pp.180-184
• /
• 2018

Gallium oxide ($Ga_2O_3$) and silicon carbide (SiC) are the material with the wide band gap ($Ga_2O_3-4.8{\sim}4.9eV$, SiC-3.3 eV). These electronic properties allow high blocking voltage. In this work, we investigated the characteristic of $Ga_2O_3$ and 4H-SiC vertical depletion-mode metal-oxide-semiconductor field-effect transistors. We demonstrated that the blocking voltage and on-resistance of vertical DMOSFET is dependent with structure. The structure of $Ga_2O_3$ and 4H-SiC vertical DMOSFET was designed by using a 2-dimensional device simulation (ATLAS, Silvaco Inc.). As a result, 4H-SiC and $Ga_2O_3$ vertical DMOSFET have similar blocking voltage ($Ga_2O_3-1380V$, SiC-1420 V) and then when gate voltage is low, $Ga_2O_3-DMOSFET$ has lower on-resistance than 4H-SiC-DMOSFET, however, when gate voltage is high, 4H-SiC-DMOSFET has lower on-resistance than $Ga_2O_3-DMOSFET$. Therefore, we concluded that the material of power device should be considered by the gate voltage.

• Journal of the Korean Applied Science and Technology
• /
• v.32 no.2
• /
• pp.330-338
• /
• 2015

The influences of fluorescence, scattering, and flocculation in turbid material by light scattering of N-propyl-N,N-dimethylethanolamine, fluorescence agent and absorption agent were interpreted for the scattered fluorescence intensity and wavelength. They have been studied the molecular properties by the spectroscopy of laser induced fluorescence (LIF) and flocculation. The effects of optical properties in scattering media have been found by the optical parameters(${\mu}_s$, ${\mu}_a$, ${\mu}_t$). Flocculation is an important step in many solid-liquid separation processes and is widely used. When two particles approach each other, interactions of several colloid particles can come into play which may have major effect on the flocculation and LIF process. The values of scattering coefficient ${\mu}_s$ are large by means of the increasing scattering of scatterer, The values have been found that the slope decays exponentially as a function of concentration from laser source to detector by our experimental result. It may also aid in designing the best model for oil chemistry, bio-pharmaceutical, laser medicine and application of medical engineering on LIF and coagulation in particle transport mode.