• Journal of Korean Society of Disaster and Security
• /
• v.14 no.1
• /
• pp.9-21
• /
• 2021

Recently, the importance of eco-friendly power generation facility using renewable energy has newly appeared. Hydropower plant is a very important source of electricity generation and supply which is very important to secure safety because it is commonly connected with multi facility and operated on a large scale. In this study, a scenario-based analysis method was suggested to assess vulnerability of a penstock system caused by water hammer commonly occurred in the operation of hydropower plants. A hypothetical hydropower plant was used to demonstrate the applicability of a transient analysis model. In order to verify reliability of the model, the prediction of pressure behaviors were compared with the results of commercial model (SIMSEN) and measured data, then a real hydroelectric power plant was applied to develop all potential water hammer scenarios during the actual operation. The scenario-based simulation and vulnerability assessment for water hammer in the penstock system were performed with internal and external load conditions. The simulation results indicated that the vulnerability of a penstock system was varied with the operating conditions of hydropower facilities and significantly affected by load combination consisting of different load scenarios. The proposed numerical method could be an useful tool for the vulnerabilityty assessment of the hydropower plants due to water hammer.

• Journal of the Korea Society of Computer and Information
• /
• v.29 no.2
• /
• pp.31-41
• /
• 2024

Deep learning models show excellent performance in tasks such as image classification and object detection in the field of computer vision, and are used in various ways in actual industrial sites. Recently, research on improving robustness has been actively conducted, along with pointing out that this deep learning model is vulnerable to hostile examples. A hostile example is an image in which small noise is added to induce misclassification, and can pose a significant threat when applying a deep learning model to a real environment. In this paper, we tried to confirm the robustness of the edge-learning classification model and the performance of the adversarial example detection model using it for adversarial examples of various algorithms. As a result of robustness experiments, the basic classification model showed about 17% accuracy for the FGSM algorithm, while the edge-learning models maintained accuracy in the 60-70% range, and the basic classification model showed accuracy in the 0-1% range for the PGD/DeepFool/CW algorithm, while the edge-learning models maintained accuracy in 80-90%. As a result of the adversarial example detection experiment, a high detection rate of 91-95% was confirmed for all algorithms of FGSM/PGD/DeepFool/CW. By presenting the possibility of defending against various hostile algorithms through this study, it is expected to improve the safety and reliability of deep learning models in various industries using computer vision.

• Journal of the Korean Geotechnical Society
• /
• v.30 no.1
• /
• pp.37-47
• /
• 2014

Unsaturated hydraulic conductivity (HC) is integrated theoretically from soil water retention curves (SWRC) by Mualem capillary model, but the prediction of HC is extremely sensitive to small variation of matric suction near saturation. Near saturation, the Mualem HC based on smooth SWRC decreases abruptly and has problems in the reliability of hydraulic behavior and the stability of numerical solutions. To improve van Genuchten-Mualem (VGM) HC, the van Genuchten SWRC model is modified within range of low matric suction (arbitrary air entry pressure). At an arbitrary air entry pressure, the VG SWRC is linearized in log scale until full saturation. The modified VG SWRC does not affect the fit of actual retention behavior and either the parameters of original VG SWRC fit. Using the modified VG SWRC, the VGM HC is modified to integrate for each interval decomposed by arbitrary air entry pressure. An analytical solution on modified VGM HC is proposed each interval, to protect the rapid change in HC near saturation. For silty soils, VGM models of HC function underestimate the unsaturated permeability characteristics and especially show rapid reduction near saturation. The modified VGM model predicts more accurate HC functions for Korean weathered soils. Furthermore, near saturation, the saturated HC is conserved by the modified VGM model. After 2-D infiltration analysis of an actual slope, the hydraulic behaviors are compared for VGM and the modified models. The prediction by the proposed model conserved the convergence of solutions on various rainfall conditions. However, the solution by VGM model did not converge since the conductivity near saturation reduced abruptly for heavy rainfall condition. Using VGM model, the factor of safety is overestimated in both initial and final stage during heavy rainfall. Stability analysis based on infiltration analysis could simulate the actual slope failure by the proposed model on HC.

• Journal of the Society of Disaster Information
• /
• v.16 no.2
• /
• pp.237-247
• /
• 2020

Purpose: This study aims to draw the problems and improvements of NFSC 103 sprinkler head discharge obstruction prevention standard to increase the fire extinguishing performance of sprinkler system in order to contribute to the protection of people's lives and property in the event of a fire. Method: NFSC 103 was compared to NFPA 13 which is the latest 2019 version in U.S in terms of sprinkler head discharge obstruction prevention. Results: This study found that NFSC 103 doesn't define even the basic concept of sprinkler discharge obstruction. And NFSC 103 doesn't have detailed standard for side wall sprinkler head discharge obstruction prevention as well as the "height" criteria of the "three times" separation rule. Conclusion: NFSC 103 needs a lot of supplements and improvements such as the addition of definition for water discharge obstructions, the adoption of sidewall sprinkler heads standard for preventing water discharge obstructions and the additional establishment of the "three times rule" considering the "height" of obstacles to promote the advance of fire safety standard equal or above fire fighting advanced country and increase the reliability for the suppression performance of sprinkler system.

• Journal of the Korean Institute of Gas
• /
• v.5 no.2 s.14
• /
• pp.30-35
• /
• 2001

Presently, working gas pipelines are being subjected to the influence of construction vibration. Especially on subway and road construction, gas pipelines are being influenced to construction vibration caused by use of construction equipment, passage of a large-sized vehicle and blasting. Buried gas pipelines are subjected to the influence of vibration caused by blast in the vicinity of pipeline, exposed gas pipelines are subjected to the influence of vehicle vibration. Therefore, in the study, it is developed to vibration prediction program of gas pipeline by analyzing measured construction vibration. This program is able to predict vibration of gas pipeline according to field conditions by using the results of structural finite element analysis and empirical equation by reliability analysis. And, this program contains the database of construction vibration. Additionally, this program is able to compute estimated blast vibration equation using measured blast vibration data in the field and to form graph of allowable charging gunpowder per delayed-action with the change of blast velocity. Therefore, field workers are able to predict construction vibration around gas pipeline and estimate safety of gas pipeline.

• Journal of the Korea Institute of Information Security & Cryptology
• /
• v.23 no.4
• /
• pp.743-755
• /
• 2013

In recent years, the malicious apps with malicious code in normal apps are increasingly redistributed in Android market, which may incur various problems such as the leakage of authentication information and transaction information and fraudulent transactions when banking apps to process the financial transactions are exposed to such attacks. Thus the financial authorities established the laws and regulations as an countermeasures against those problems and domestic banks provide the integrity verification functions in their banking apps, yet its reliability has not been verified because the studies of the safety of the corresponding functions have seldom been conducted. Thus this study suggests the vulnerabilities of the integrity verification functions of banking apps by using Android reverse engineering analysis techniques. In case the suggested vulnerabilities are exploited, the integrity verification functions of banking apps are likely to be bypassed, which will facilitate malicious code inserting attacks through repackaging and its risk is very high as proved in a test of this study. Furthermore this study suggests the specific solutions to those vulnerabilities, which will contribute to improving the security level of smartphone financial transaction environment against the application forgery attacks.

• Journal of the Korean Society for Aeronautical & Space Sciences
• /
• v.42 no.8
• /
• pp.640-647
• /
• 2014

Infrared signature of aircraft exhaust plume is the critical factor for aircraft survivability. To improve the military aircraft survivability, the accurate prediction of infrared signature for the propulsion system is needed. The numerical analysis of thermal fluid field for nozzle inflow, free stream flow, and plume region is conducted by using the in-house code. Weighted Sum of Gray Gases Model based on Narrow Band with regrouping is adopted to calculate the spectral infrared signature emitted from aircraft exhaust plume. The accuracy and reliability of the developed code are validated in the one-dimensional band model. It is found that the infrared radiant intensity is relatively more strong in the plume through the analysis, the results show the different characteristic of the spectral infrared signature along the temperature, the partial pressure, and the species distribution. The continuous spectral radiant intensity is shown near the nozzle exit due to the emission from the nozzle wall.

• The Journal of the Korea institute of electronic communication sciences
• /
• v.15 no.3
• /
• pp.421-432
• /
• 2020

This paper proposes a 2-kW solid-state power amplifier (SSPA) developed by employing power amplifier pallets designed using gallium-nitride high electron mobility transistors, which is used in S-band military radars and to replace existing traveling-wave tube amplifier (TWTA). The SSPA consists of a high-power amplifier module, which combines eight power amplifier pallets, a drive amplifier module, a digital control module, and a power supply unit. First, the amplifier module and component were integrated into a small package to account for space limitations; next, an on-board harmonic filter was fabricated to reject spurious components; and finally, an auto gain control system was designed for various duty ratios because recent military radar systems are all active phase radars using the pulse operation mode. The developed SSPA exhibited a max gain of 48 dB and an output power ranging between 63-63.6 dBm at a frequency band of 3.1 to 3.5 GHz. The auto gain control function showed that the output power is regulated around 63 dBm despite the fluctuation of the input power from 15-20 dBm. Finally, reliability of the developed system was verified through a temperature environment test for nine hours at high (55 ℃) / low (-40℃) temperature profile in accordance with military standard 810. The developed SSPA show better performance such as light weight, high output, high gain, various safety function, low repair cost and short repair time than existing TWTA.

• International Journal of Highway Engineering
• /
• v.12 no.2
• /
• pp.9-16
• /
• 2010

Roughness is the most important factor to maintain the road performance, and affects greatly on the design life in Jointed Plain Concrete pavements. Also, the factors the evaluate pavement‘s commonality is the three method such as functionality, safety and structural performance. In evaluating function of road, representative factors is the roughness, which has been used to determine maintenance time as key standard. As research for roughness is absence in pavement design. Applied roughness-model had a low-reliability in Korea. Therefore, it is needed to develop reliable model in road roughness. In this research, uniform specific is applied to distribute them after selecting the concrete pavements. Concrete pavement is divided by sections of 238. Total length of this sections has 281km and account for 16% of total road length in korean concrete pavements for selected sections. Considering the korean roughness-model, the evaluation of roughness is performed for the freezing index, average annual rainfall, condition for the base, the amount of traffic as well as spalling(%), cracking(%), age(year) at the selected section at the selected section. Also, additional sections is selected to evaluate various age which affects on the roughness. As a result of the analysis, it showed that spalling(%), cracking(%), age(year), and the condition of the base affected road roughness. When the correlation with the road roughness was analyzed, the reliable model for road roughness was proposed, and the ratio that can explain road roughness was R2-68.8% and P value-0 which is statistically meaningful.

• Journal of Korean Society of Transportation
• /
• v.33 no.6
• /
• pp.554-563
• /
• 2015

This study dealt with developing an accident model for rural signalized intersections with random parameter negative binomial method. The limitation of previous count models(especially, Poisson/Negative Binomial model) is not to explain the integrated variations in terms of time and the distinctive characters a specific point/segment has. This drawback of the traditional count models results in the underestimation of the standard error(t-value inflation) of the derived coefficient and finally affects the low-reliability of the whole model. To solve this problem, this study improves the limitation of traditional count models by suggesting the use of random parameter which takes account of heterogeneity of each point/segment. Through the analyses, it was found that the increase of traffic flow and pedestrian facilities on minor streets had positive effects on the increase of traffic accidents. Left turning lanes and median on major streets reduced the number of accidents. The analysis results show that the random parameter modeling is an effective method for investigating the influence on traffic accident from road geometries. However, this study could not analyze the effects of sequential changes of driving conditions including geometries and safety facilities.