• Journal of the Society of Disaster Information
• /
• v.17 no.3
• /
• pp.608-616
• /
• 2021

Purpose: Safety inspections by existing personnel have been limited in evaluation and data securing due to concerns about the safety of technicians or difficulty in accessing them, and are becoming a bigger problem as the number of maintenance targets increases due to the aging of facilities. As drone technology develops, it is possible to ensure the safety of personnel, secure visual data, and diagnose quickly, and use it is increasing as safety inspection of facilities by drones was introduced recently. In order to further enhance utilization, it is considered necessary to base a consideration standard for facility appearance investigation by drones, and in this paper, research was conducted on dams. Method: To calculate the quality, existing domestic safety inspection and drone-related consideration standards were investigated, and procedures related to safety inspection using drones were compared and analyzed to review work procedures and construction types. In addition, empirical data were collected through drone photography and elevation image production for the actual dam. Result: Work types for safety inspection of facilities using drones were derived, and empirical survey results were collected for two dams according to work types. The existing guidelines were applied for the adjustment ratios for each structural type and standard of the facility, and if a meteorological reference point survey was necessary, the unmanned aerial vehicle survey of the construction work standard was applied. Conclusion: The finer the GSD in appearance investigation using drones, the greater the number of photographs taken, and the concept of adjustment cost was applied as a correction to calculate the consideration standard. In addition, it was found that the problem of maximum GSD indicating limitations should be considered in order to maintain the safe distance.

• Journal of Internet of Things and Convergence
• /
• v.8 no.6
• /
• pp.55-63
• /
• 2022

In recent years, with the rapid development of large and medium-sized urban rail transit in China, the total operating mileage of high-speed railway and the total number of EMUs(Electric Multiple Units) are rising. The system complexity of high-speed EMU is constantly increasing, which puts forward higher requirements for the safety of equipment and the efficiency of maintenance.At present, the maintenance mode of high-speed EMU in China still adopts the post maintenance method based on planned maintenance and fault maintenance, which leads to insufficient or excessive maintenance, reduces the efficiency of equipment fault handling, and increases the maintenance cost. Based on the intelligent operation and maintenance technology of PHM(prognostics and health management). This thesis builds an integrated PHM platform of "vehicle system-communication system-ground system" by integrating multi-source heterogeneous data of different scenarios of high-speed EMU, and combines the equipment fault mechanism with artificial intelligence algorithms to build a fault prediction model for traction motors of high-speed EMU.Reliable fault prediction and accurate maintenance shall be carried out in advance to ensure safe and efficient operation of high-speed EMU.

• Journal of the Korean Association of Geographic Information Studies
• /
• v.24 no.4
• /
• pp.194-208
• /
• 2021

The Hwanggang Dam in North Korea is located upstream of the Imjin River which is a shared river in the border area. It is known to have a reservoir capacity of 350 million cubic meters and releases a discharge primarily for generating hydroelectric power and partly for transferring to the Yesung River basin. Due to the supply of water from the Hwanggang Dam to another basin, the flow of the Imjin River has decreased, which has a negative impact on the water supply, river maintenance flow, water quality, and ecological environment in Korea. However, due to the special national security issue of the South and North Korea border region, the hydrological data is not shared, and the operation method of the Hwanggang Dam is unknown, so there is a risk of damage to the southern part of the downstream area. In this study, the monthly diversion as the long-term runoff concept was derived through the calibrated hydrological model based on optical remotely sensed Images and water balance analysis. As a result of the water balance analysis from January 2019 to September 2021, the average diversion of the Hwanggang Dam was 29.2m³/s, which is equivalent to 922 million tons per year and 45.6% of the annual inflow of 2.02 million tons into the Hwanggang Dam.

• KIPS Transactions on Software and Data Engineering
• /
• v.11 no.4
• /
• pp.163-168
• /
• 2022

As cloud computing and the Internet of things are getting popular, the number of devices in the Internet of things computing environments is increasing. In addition, there exist various Internet-based applications, such as home automation and healthcare. In turn, existing studies explored the quality of service, such as downtime and reliability of tasks for Internet of things applications. To enhance the quality of service of Internet of things applications, cloud-fog computing (combining cloud computing and edge computing) can be used for offloading burdens from the central cloud server to edge servers. However, when devices inherit the mobility property, continuity and the quality of service of Internet of things applications can be reduced. In this paper, we propose a resource management scheme based on live migrations for mobility support in edge-based fog computing environments. The proposed resource management algorithm is based on the mobility direction and pace to predict the expected position, and migrates tasks to the target edge server. The performance results show that our proposed resource management algorithm improves the reliability of tasks and reduces downtime of services.

• The Journal of Korea Institute of Information, Electronics, and Communication Technology
• /
• v.15 no.3
• /
• pp.198-204
• /
• 2022

In this paper, the 5kW and 50kW vertical axis wind turbines were studied using the ANSYS flow analysis simulation program. The 5 kW vertical shaft wind turbine has 30 units of the number of main blades and sub-blades and the electrical characteristics were analyzed by changing the tip speed ratio (TSR) from 0.2 to 06. A 50kW vertical axis wind turbine was designed based on the electrical characteristics of a 5kW vertical axis wind turbine. When the tip speed ratio was 0.5, the 5 kW wind power generation showed the maximum output of 9.5 kW and the efficiency of 0.28. The calculation of the power current(Ip) and the power voltage(Ep) show that, as the tip speed ratio increases, the power current(Ip) decreases and the power voltage(Ep) increases. And even if the tip speed ratio was changed, 5kW wind power generation was measured for output of 5 kW or higher. When the tip speed ratio was changed from 0.3 to 0.6, 50 kW wind power generation was output more than 50 kW. When the tip speed ratio of 50kW wind power generation was 0.4, the output was 58.37 [kW] and the efficiency was 0.318, and it was confirmed that the proposed 50kW wind power generation satisfies the design conditions.

• Journal of the Korea Society for Simulation
• /
• v.31 no.1
• /
• pp.11-18
• /
• 2022

Recently, research to classify human activity using imagery has been actively conducted for the purpose of crime prevention and facility safety in public places and facilities. In order to improve the performance of human activity classification, most studies have applied deep learning based-transfer learning. However, despite the increase in the number of backbone network models that are the basis of deep learning as well as the diversification of architectures, research on finding a backbone network model suitable for the purpose of operation is insufficient due to the atmosphere of using a certain model. Thus, this study applies the transfer learning into recently developed deep learning backborn network models to build an intelligent system that classifies human activity using imagery. For this, 12 types of active and high-contact human activities based on sports, not basic human behaviors, were determined and 7,200 images were collected. After 20 epochs of transfer learning were equally applied to five backbone network models, we quantitatively analyzed them to find the best backbone network model for human activity classification in terms of learning process and resultant performance. As a result, XceptionNet model demonstrated 0.99 and 0.91 in training and validation accuracy, 0.96 and 0.91 in Top 2 accuracy and average precision, 1,566 sec in train process time and 260.4MB in model memory size. It was confirmed that the performance of XceptionNet was higher than that of other models.

• Journal of the Society of Disaster Information
• /
• v.18 no.1
• /
• pp.51-62
• /
• 2022

Purpose: This study analyzes accessibility of outdoor evacuation places for earthquake and the accessibility improvement effects when expanding the evacuation places in accessibility-deficient areas. In order to consider real-world evacuees, the accessibility analysis is based on service population not on resident population. Method: Location-allocation model as a GIS-based spatial optimization mode is used to analyze accessibility and vulnerable areas to evacuation places. Of location-allocation problem types, 'Maximize Coverage' method is chosen to allocate as many potential evacuees as possible to evacuation places. And impedence cutoffs or evacuation distances (times) are applied to three classes: 500m (7.5 minutes), 1,000m (15 minutes), and 1,500m (22.5 minutes). Case study area is Jung-gu areas, Seoul as a high-density downtown area. Result: Results show that accessibility-deficient areas and population to evacuation places are much more in service population than in resident population. Accessibility is significantly improved when increases when expanding the evacuation places in accessibility-deficient areas. Yet, accessibility-deficient areas are still remained since available lands are insufficient in the high-density downtown area. Conclusion: The study suggests that temporary evacuation facilities like outdoor evacuation places for earthquake need to consider real potential evacuees based not only on resident population but also on service population. Also, policy measures to provide emergency shelters need to more utilize spatial optimization tools like location-allocation model.

• KIPS Transactions on Software and Data Engineering
• /
• v.11 no.7
• /
• pp.291-298
• /
• 2022

In this paper, we present a method for dependency parsing with chunking in Korean. Dependency parsing is a task of determining a governor of every word in a sentence. In general, we used to determine the syntactic governor in Korean and should transform the syntactic structure into semantic structure for further processing like semantic analysis in natural language processing. There is a notorious problem to determine whether syntactic or semantic governor. For example, the syntactic governor of the word "먹고 (eat)" in the sentence "밥을 먹고 싶다 (would like to eat)" is "싶다 (would like to)", which is an auxiliary verb and therefore can not be a semantic governor. In order to mitigate this somewhat, we propose a Korean dependency parsing after chunking, which is a process of segmenting a sentence into constituents. A constituent is a word or a group of words that function as a single unit within a dependency structure and is called a chunk in this paper. Compared to traditional dependency parsing, there are some advantage of the proposed method: (1) The number of input units in parsing can be reduced and then the parsing speed could be faster. (2) The effectiveness of parsing can be improved by considering the relation between two head words in chunks. Through experiments for Sejong dependency corpus, we have shown that the USA and LAS of the proposed method are 86.48% and 84.56%, respectively and the number of input units is reduced by about 22%p.

• Journal of the Society of Disaster Information
• /
• v.18 no.3
• /
• pp.574-580
• /
• 2022

Purpose: As corn oil is semi-drying oil, it has more double bonds than non-drying oil and is easier to combine with oxygen. In addition, among the causes of spontaneous ignition, accidents caused by oil-soaked cloths due to oxidative heat are gradually increasing. Therefore, the purpose of this study is to understand the characteristics of spontaneous combustion according to the number of towels and the amount of corn oil at 65℃. Method: After setting the test temperature to 65℃, 25ml, 50ml, 75ml of corn oil per towel was sprayed. The central temperature of the sample rises above the set temperature. It was determined, and when the central temperature of the sample became similar to the set temperature, it was determined as non-igniting. Result: After evenly distributing 25ml of corn oil per towel, as a result of the experiment, 5 towels did not ignite, and 10 and 15 towels ignited. Also, as a result of an experiment using 50ml and 75ml of corn oil per towel, spontaneous ignition occurred when the number of towels was 5, 10, or 15 sheets. Conclusion: Even a small amount can cause a fire if the conditions for spontaneous ignition are met.

• Journal of the Korean Society of Radiology
• /
• v.16 no.5
• /
• pp.661-672
• /
• 2022

After the outbreak of the SARS-CoV2 virus that causes COVID-19, it spreads around the world with the number of infections and deaths rising rapidly caused a shortage of medical resources. As a way to solve this problem, chest X-ray diagnosis using Artificial Intelligence(AI) received attention as a primary diagnostic method. The purpose of this study is to comprehensively analyze the detection of COVID-19 via AI. To achieve this purpose, 292 studies were collected through a series of Classification methods. Based on these data, performance measurement information including Accuracy, Precision, Area Under Cover(AUC), Sensitivity, Specificity, F1-score, Recall, K-fold, Architecture and Class were analyzed. As a result, the average Accuracy, Precision, AUC, Sensitivity and Specificity were achieved as 95.2%, 94.81%, 94.01%, 93.5%, and 93.92%, respectively. Although the performance measurement information on a year-on-year basis gradually increased, furthermore, we conducted a study on the rate of change according to the number of Class and image data, the ratio of use of Architecture and about the K-fold. Currently, diagnosis of COVID-19 using AI has several problems to be used independently, however, it is expected that it will be sufficient to be used as a doctor's assistant.