• Journal of the Korean Association of Geographic Information Studies
• /
• v.2 no.2
• /
• pp.69-78
• /
• 1999

In this paper, a neural network approach to forecast Korean regional precipitation is presented. We first analyze the characteristics of the conventional models for time series prediction, and then propose a new model and its learning method for the precipitation forecast. The proposed model is a layered network in which the outputs of a layer are buffered within a given period time and then fed fully connected to the upper layer. This study adopted the dual connections between two layers for the model. The network behavior and learning algorithm for the model are also described. The dual connection structure plays the role of the bias of the ordinary Multi-Layer Perceptron(MLP), and reflects the relationships among the features effectively. From these advantageous features, the model provides the learning efficiency in comparison with the FIR network, which is the most popular model for time series prediction. We have applied the model to the monthly and seasonal forecast of precipitation. The precipitation data and SST(Sea Surface Temperature) data for several decades are used as the learning pattern for the neural network predictor. The experimental results have shown the validity of the proposed model.

• Korean journal of applied entomology
• /
• v.38 no.2
• /
• pp.123-128
• /
• 1999

Recently, entomopathogenic nematodes have received a considerable attention as biologicalcontrol agents. For in vitro cultivation, storage and transportation of nematodes, oxygen supply isextremely impotant due to its limited solubility and mass transfer problem. The oxygen uptake rates(OURs) of four different Steinernema species were measured in a 5L bioreactor at varying temperatures.The OURs of the Steinernema spp. were below 0.5 x mmolO'||'&'||' . min in the range of 13-17$^{\circ}$C. TheOURs (mmo102/L - min) of S. glaseri Dongrae and S. carpocapsae Pocheon strains were 0.4 x lo-', 0.75x lo-\ulcorner at 21$^{\circ}$C, 1.5 x lo-\ulcorner, 3.2 x 10-2 at 25"C, and 2.8 x lo-', 5.8 x lo-\ulcorner at 29"C, respectively. However,the OURs were not significantly altered by the agitation speed of 50-150 rpm. The specific oxygenuptake rates (qol) of S. glaseri NC, S. glaseri Dongrae, S. glaseri Mungyeong and S. carpocapsaePocheon strains were 0.3 x 0.5 x 0.3 x and 0.2 x mmolO~/cell min at 25"C,respectively. As the nematode size and temperature were increased, the qo, was also increased.the qo, was also increased.

• Journal of Adhesion and Interface
• /
• v.16 no.3
• /
• pp.101-107
• /
• 2015

Epoxy resin toughening agents such as core/shell nanoparticles, CTBN epoxy, polyester polyols, and polyurethane have been widely used in order to compensate for the brittleness and improve the impact resistance of the epoxy resin. In this work, a few tougheners mentioned above were individually added into adhesive compositions to observe the effects of physical and mechanical properties. Both flexural strength and flexural modulus were measured with UTM while impact strength was analyzed with Izod impact tester. The obtained results showed that the addition of toughening agents afforded positive performance in terms of flexibility and impact resistance of the cured epoxy resin. Furthermore, DMA experiments suggested that the trends of storage modulus data of each epoxy resin composition coincided with the trends of flexural modulus data. FE-SEM images showed that toughening agents formed circled-shape particles when it was cured in epoxy resin composition at high temperature by phase separation. The existence of particles in the cured samples explains why epoxy resin with toughener has higher impact resistance.

• Korean Journal of Remote Sensing
• /
• v.35 no.6_3
• /
• pp.1187-1195
• /
• 2019

We, for the first time, developed a Raman lidar system which can remotely detect surface CO₂ volume mixing ratio (VMR). The Raman lidar system consists of the Nd: YAG laser of wavelength 355 nm with 80 mJ, an optical receiver, and detectors. Indoor CO₂ cell measurements show that the accuracy of the Raman lidar system is calculated to be 99.89%. We carried out the field measurement using our Raman lidar at Pukyong National University over a seven-day period in October 2019. The results show good agreement between CO₂ VMRs measured by the Raman lidar (CO₂ Raman Lidar) and those measured by in situ instruments (CO₂ In situ) which located 300 m and 350 m away from the Raman lidar system. The correlation coefficient (R), mean absolute error (MAE), and root mean square error (RMSE) between CO₂ In situ and CO₂ Raman Lidar are 0.67, 2.78 ppm, and 3.26 ppm, respectively.

• The Journal of The Korea Institute of Intelligent Transport Systems
• /
• v.15 no.1
• /
• pp.66-76
• /
• 2016

Determination of type of a vehicle is being used in various areas such as collecting tolls, collecting statistical traffic data and traffic prognosis. Because most of the vehicle type classification systems depend on vehicle length indirectly or directly, highly reliable automatic vehicle length measurement system is crucial for them. This study makes use of a pencil beam laser rangemeter and devises a mechanical device which rotates the laser rangemeter. The implemented system measures the range between a point and the laser rangemeter then indicates it as a spherical coordinate. We obtain several silhouettes of cross section of the vehicle, the rate of change of the silhouettes, signs of the rates then squares the rates to apply cell averaging constant false alarm rate (CA-CFAR) technique to find out where the border is between the vehicle and the background. Using the border and trigonometry, we calculated the length of the vehicle and confirmed that the calculated vehicle length is about 94% of actual length.

• Journal of Korea Spatial Information System Society
• /
• v.6 no.1 s.11
• /
• pp.31-50
• /
• 2004

With the growth of wireless communication networks and mobile devices taking in GPS, Location-Based Service(LBS) is becoming an integral part of mobile applications. LBS can deal with location-aware features such as persons holding mobile phones or vehicles equipped with GPS, and provide the users with the location information of the features. Thus it is necessary to develop moving object database systems to store, manage, and query moving objects which change their locations continuously as time passes. In this paper, we design and implement a query processing component which deals with moving objects as a key data type. For this component, we define a new SQL-like query language(called MOQL) and as a consequence, design and implement modules that analyze and execute queries. It supports various types of operators that process range queries, infer topological relations, compute trajectories, and find k-nearest neighbors. It can be used as a subsystem if other application systems which deal moving objects and also supports ADO.NET interface that can be used to interact end-users.

• Proceedings of the Korea Database Society Conference
• /
• 1999.06a
• /
• pp.335-345
• /
• 1999

오늘날 날로 증가하는 물류비는 개별 기업은 물론 국가 전체의 수출 경쟁력을 약화시키는 주요 원인으로 지적되고 있다. 그러나 그동안 우리나라에서는 물류비 절감을 위한 종합적이고 체계적인 대책이 이루어지지 못하였다. 특히 본 논문의 연구대상인 육상물류의 경우 그 비중이 전체 화물 운송의 60% 이상을 차지함에도 불구하고 심각한 교통체증 및 물류기반 시설의 미비, 효율적인 정보시스템의 미비 등으로 인하여 물류비가 계속 증가하는 양상을 보여 왔다. 따라서 본 논문에서는 우리나라 육상물류시스템이 안고 있는 문제점의 해결을 위한 방안들 중의 하나로 정보기술의 활용에 관한 내용을 다루고 있다. 즉 영세한 기업들도 누구나 손쉽게 이용할 수 있도록 인터넷을 이용한 육상물류중개시스템의 개발에 관한 내용을 소개하고 있다. 육상물류중개시스템은 복합화물주선업체인 (주) 대형물류와 함께 개발한 시스템으로 인터넷을 통하여 화주의 화물 운송의뢰를 접수받아 이를 여러 운송업체에게 제공해주는 역할을 수행하게 된다. 특히 육상물류중개시스템은 화물의 운송과 관련하여 발생하는 다양한 정보들을 데이터베이스에 저장하여 두었다가 세관을 비롯한 터미날에 대한 각종 신고업무에 이용할 수 있으며, 이밖에도 교통정보 및 화물 위치정보 등 다양한 서비스를 제공해줄 수 있다. 따라서 운송업체의 공차율을 줄이고 화주에게는 자신의 화물에 대한 정보를 실시간으로 전달해 줄 수 있다는 장점이 있다. 또한 이러한 육상물류중개시스템은 현재 개발중인 통합데이터베이스를 기반으로 한 항만물류원스톱서비스 시스템과 연계되어 차후에는 물류원스톱시스템으로 발전할 수 있을 것이다. 연구가 진행되고 있는 인공신경망과의 모형결합을 통해 기존연구와는 다른 새로운 통합예측방법론을 제시하고자 한다. 본 연구에서 제시하는 통합방법론은 크게 2단계 과정을 거쳐 예측모형으로 완성이 된다. 즉, 1차 모형단계에서 원시 재무시계열은 먼저 웨이블릿분석을 통해서 노이즈가 필터링 되는 동시에, 과거 재무시계열의 프랙탈 구조, 즉 비선형적인 움직임을 보다 잘 반영시켜 주는 다차원 주기요소를 가지는 시계열로 분해, 생성되며, 이렇게 주기에 따라 장단기로 분할된 시계열들은 2차 모형단계에서 신경망의 새로운 입력변수로서 사용되어 최종적인 인공 신경망모델을 구축하는 데 반영된다.ocioeconomic impacts are resulted from the program. It would be useful for the means of (ⅰ) fulfillment of public accountability to legitimate the program and to reveal the expenditure of pubic fund, and (ⅱ) managemental and strategical learning to give information necessary to improve the making. program and policy decision making, The objectives of the study are to develop the methodology of modeling the socioeconomic evaluation, and build up the practical socioeconomic ev

• KOREAN JOURNAL OF CROP SCIENCE
• /
• v.49 no.3
• /
• pp.227-231
• /
• 2004

A total of 219 polymerase chain reaction tests of genetically modified (GM) DNA sequences in soybean seeds and soybean sprouts were conducted during 2000-2001. No CM gene was found in 96 tests of soybean seeds. However, either a functional CP4EPSPS gene or the 355 promoter gene was found three times in 2000 and eight times in 2001, in between 0.01 and 0.17％ of soybean spout products, in 123 tests. Since the amount of GM genes was much less than the threshold limit of 3％, none of the 11 positive soybean-sprout samples needed to be libeled GM crops. Of these, seven sprout samples were from domestic seeds and four were from seeds imported from China. To find the contamination route, the raw materials, seed surface, floor of the storage room, area around the selection machine, surface of the packaging film and corn powder used in the package were tested. The 35S promoter gene was detected in only two samples of the corn powder (0.1％). Although we could not find the cause of the GM contamination, the sprout package film is one possibility. In total,8.9％ of the soybean sprout tests were GM positive, but the amounts were much less than the threshold of 3％. This means that there are frequent false-positives and these would threaten the sprout industry if GMO were decided qualitatively. Food companies should make their safety data available to the public and make an effort to address people's concerns about GM food more openly. In addition, there is a need to establish a quantitative test for GM genes in sampled water and a sampling method for raw materials.

• Journal of Food Hygiene and Safety
• /
• v.22 no.3
• /
• pp.151-158
• /
• 2007

Among 68 strains of lactic acid bacteria (LAB) isolated from Dongchimi, a strain K11 was selected due to its bactericidal activity against Escherichia coli O157 The strain K11 was identified as Lactobacillus plantarum, based on physiological and biochemical characteristics. In the late exponential phase, La. plantarum K11 showed maximum bacteriocin activity (12,800 BU/mL) and maintained until the early stationary phase. The bacteriocin activity was completely inactivated by all the proteolytic enzymes such as pepsin, protease, proteinase K, papain, chymotrypsin, and trypsin, but the activity was not affected by catalase, a-amylase, lysozyme, and lipase, suggesting proteinaceous nature of the bacteriocin. Additionally, this activity was not affected in the pH range from 3.0 to 9.0 and under storage conditions like 30 days at -20,4, or $25^{\circ}C$. Although the bacteriocin activity was absolutely lost after 15 min treatment at 121, it was relatively stable at $70^{\circ}C$ for 60 min or $100^{\circ}C$ for 30 min. The activity was disappeared by treatment with acetone, benzene, ethanol, or methanol, but it was not affected by treatment with chloroform or hexane. The antibacterial activity of the bacteriocin was good against some LAB including Lactobacillus spp., Enterococcus spp., and Streptococcus spp., but not against food-borne pathogens such as Bacillus spp., Listeria spp., and Staphylococcus spp. as well as yeasts and molds. Especially, some intestinal bacteria such as Enterobacter aerogenes and E. coli were significantly affected by the bacteriocin of La, plantarum K11. Furthermore, the addition of 640 BU/mL resulted in the complete clearance of E. coli O157 after 10 hr.

• Journal of Environmental Science International
• /
• v.12 no.4
• /
• pp.487-496
• /
• 2003

The purpose of this study was to develop the preliminary source fingerprints of volatile organic compounds (VOC). The source categories studied were vehicles, gasoline vapor, gasoline storage tank, coating, dry cleaning and road covering. The source samples were collected using 6L electro-polished stainless steel canisters for about 20 seconds. From this study, the main component emitted from VOC sources in Korea was toluene. The toluene proportion for road covering, vehicles, coating and gasoline vapor were 35, 18, 16 and 5%, respec- tively. The C$_2$-C$\sub$5/ alkane and alkene compounds were mainly emitted from vehicles, gasoline vapor and gasoline storage tank. The main compounds of coating were m/p-xylene(34%), toluene(16%), 1,2,4-TMB(10%) and o-xylene(9%), which are aromatic hydrocarbons. In the case of dry cleaning, nonane(41%), 1,2,4-TMB (22%) and 1,3,5-TMB(13%) were mainly emitted.