• The Transactions of the Korea Information Processing Society
• /
• v.2 no.2
• /
• pp.148-155
• /
• 1995

When we analize many amount of patterns, it is necessary for these patterns are to be clustering into several groups according to a certain evaluation function. This process, in case that there are lots of input patterns, needs a considerable amount of computations and is reqired parallel algorithm for these. To solve this problem, this paper propose parallel clustering algorithm which parallelized k-means algorithm and implemented it under the MIMD parallel computer based message passing. The result is through the experiment and performance analysis, that this parallel algorithm is appropriate in case these are many input patterns.

• International Journal of Highway Engineering
• /
• v.15 no.1
• /
• pp.47-56
• /
• 2013

PURPOSES: This study is to evaluate a ground water level effect on frost heaving in road pavements. METHODS: The effects of water table on frost heaving in pavement systems were evaluated from the mechanical analysis using FROST program. The input parameters and boundary conditions were determined by considering climates, pavement sections, and material properties specially subgrade soil types in Korea. RESULTS: When the water table located above the freezing depth, amount of frost heaving caused by freezing the water in pavement itself was big enough to damage in pavement system, although pavement system consists of fully non-frost-susceptible materials with sufficient thickness of anti-freezing layer. The amount of frost heaving was decreased rapidly with increasing the distance between the water table and freezing depth. CONCLUSIONS: It was concluded that there is no engineering problems related with frost heaving in practical sense when the distance between freezing depth and water table is over 1.5m for having subgrade soils less than 50% of #200 sieve passing to meet specification on quality control in Korea.

• Journal of the Korean Society of Manufacturing Process Engineers
• /
• v.20 no.7
• /
• pp.113-120
• /
• 2021

An experimental device was designed to control the opening of a damper via operating the folding blade drive of the device and to control the amount of air flowing through the damper. In addition, an inverter was installed in the blower to control its fan rotation speed and hence the amount of air flowing through the damper. An experimental study was conducted on the opening of the folding blade damper and changes in the rotational speed of the blower. From the results, the theoretical air volume of the folding blade damper and experimental air volume were observed to be in good agreement within an error range of ±3%. As the mass flow rate of the air passing through the folding blade damper increases proportionally with the changes in damper opening and fan rotation speed, the performance of the damper can be controlled proportionally. The mass flow rate was also observed to increase linearly; therefore, the mass flow rate of the air passing through the folding blade damper increases proportionally with changes in the rotation speed of the blower, such that the performance of the damper is proportional to a constant air volume even with varying rotation speeds of the blower.

• Korean Journal of Applied Biomechanics
• /
• v.20 no.3
• /
• pp.319-325
• /
• 2010

The purpose of this research was to investigate the differences in the ski carving turn motion according to level of expertise. The posture and movement of 6 skiers nearby the fall-line was evaluated with a biomechanical approach focusing the rotational mechanics. The slope was at an angle of $9^{\circ}$ and the following variables were measured and calculated: tangential velocity, change of COM height after passing fall-line, width between feet, angle between upper body and thigh, trunk angle, average radius of curvature and average centripetal force. The expert skiers minimized their center of mass height movement and maintained the width of between their feet after the passing the fall-line in comparison with the beginners and intermediate skiers. The experts restrained themselves from pushing their upper body downward after the turn to maximize the centripetal force. The experts in comparison with the beginners and intermediate skiers during the turn didn't have to reduce their radius of curvature to maintain a high centripetal force. It was concluded, that the most important factor affecting the centripetal force, was for the beginners and intermediate skiers, to minimize their movement while using the appropriate amount of edging.

• Asian-Australasian Journal of Animal Sciences
• /
• v.3 no.1
• /
• pp.39-46
• /
• 1990

This experiment was conducted to study the effects of blood-mixed and heat-treated protein feeds on protein degradation in the rumen, flow of protein to the abomasums and availability of undegraded protein in the intestine of sheep in a $4{\times}4$ Latin square design. Soybean oil meal, rapeseed meal, and whole soybean were mixed with fresh swine blood and dried at $140^{\circ}C$ for 2 h. Proportionate disappearance of apparently digested OM in the postrumen for the blood and heat treated protein group was ranged from 43.2 to 50.5% as compared with 28.0% for the unheated soybean oil meal diet. The treated protein supplements were resulted in greater total N and NAN flow passing at the abomasums than untreated soybean oil meal diet was fed. The quantities of undegraded feed N passing at the abomasums for the treated protein diets was approximately twice as high as that of the untreated soybean oil meal diet and the estimated amount of undegraded N of the protein supplement itself was 79.1 to 84.2% as compared with 15% of soybean oil meal.

• The Journal of the Korea Contents Association
• /
• v.16 no.8
• /
• pp.419-426
• /
• 2016

The research is done to analyze the change of personal dosimeter according to the elapsed times(24 hours, 1 week, 2 weeks, 3 weeks, 4 weeks) and magnetic field and find out the effective exposure treatment for radiation workers. At first, research the heat treatment and radiation of grouped TLD and keep them in different environments-exposed separately to observe the consequences of glow curve and the level of radiation exposure. As a result, we could find that 24 hours passing TLD group showed the difference in glow curve and the level of radiation. This can be considered as the change caused by magnetic exposure. Also the average radiation exposure level of TLD group, unexposed to the magnetic field, was 15.41 mSv. And the average radiation exposure level of TLD group, exposed to the magnetic field, was 14.83 mSv which decreased the biggest amount(3.80%) among the other groups. If a radiation worker, who works in PET-MRI room, uses TLD as a personal dosimeter, the level of real radiation exposure caused by exposure to the magnetic field won't change significantly as recorded at a regular record cycle but with not regular record but interim record, the lower exposure dose will be appeared than the real level of radiation.

• Phonetics and Speech Sciences
• /
• v.10 no.1
• /
• pp.39-48
• /
• 2018

A typical statistical parametric speech synthesis (text-to-speech, TTS) system consists of separate modules, such as a text analysis module, an acoustic modeling module, and a speech synthesis module. This causes two problems: 1) expert knowledge of each module is required, and 2) errors generated in each module accumulate passing through each module. An end-to-end TTS system could avoid such problems by synthesizing voice signals directly from an input string. In this study, we implemented an end-to-end Korean TTS system using Google's Tacotron, which is an end-to-end TTS system based on a sequence-to-sequence model with attention mechanism. We used 4392 utterances spoken by a Korean female speaker, an amount that corresponds to 37% of the dataset Google used for training Tacotron. Our system obtained mean opinion score (MOS) 2.98 and degradation mean opinion score (DMOS) 3.25. We will discuss the factors which affected training of the system. Experiments demonstrate that the post-processing network needs to be designed considering output language and input characters and that according to the amount of training data, the maximum value of n for n-grams modeled by the encoder should be small enough.

• Proceedings of the KSRS Conference
• /
• v.1
• /
• pp.402-405
• /
• 2006

Radiometric calibration of optical image data is necessary to convert raw digital number (DN) value of each pixel into a physically meaningful measurement (radiance). To extract rather quantitative information regarding biophysical characteristics of the earth surface materials, radiometric calibration is often essential procedure. A sensor detects the radiation of sunlight interacted atmospheric constituents. Therefore, the amount of the energy reaching at the sensor is quite different from the initial amount reflected from the surface. To achieve the target reflectance after atmospheric correct, an initial step is to convert DN value to at-sensor radiance. A linear model, the simplest radiometric model, is applied to averaged spectral radiance for this conversion. This study purposes to analyze the sensitivity of several factors affecting on radiance for carrying out absolute radiometric calibration of panchromatic images from KOMPSAT2 launched at July, 2006. MODTRAN is used to calculate radiance at sensor and reflectance of target is measured by a portable spectro-radiometer at the same time the satellite is passing the target for the radiometric calibration. As using different contents of materials composing of atmosphere, the differences of radiance are investigated. Because the spectral sensitivity of panchromatic images of KOMPSAT2 ranges from 500 to 900 nm, the materials causing scattering in visible range are mainly considered to analyze the sensitivity. According to the verified sensitivity, direct measurement can be recommenced for absolute radiometric calibration.

• Advances in concrete construction
• /
• v.9 no.4
• /
• pp.375-386
• /
• 2020

In the present study, the effect of basalt, glass, and hybrid glass-basalt fibers on mechanical properties and fracture behavior of self-compacting concrete (SCC) mixes have been assessed at room and elevated temperatures. To do so, twelve mix compositions have been prepared such that the proper workability, flowability, and passing ability have been achieved. Besides, to make comparison possible, water to binder ratio and the amount of solid contents were kept constant. Four fiber dosages of 0.5, 1, 1.5, and 2% (by concrete volume) were considered for monotype fiber reinforced mixes, while the total amount of fiber were kept 1% for hybrid fiber reinforced mixes. Three different portions of glass and basalt fiber were considered for hybridization of fibers to show the best cocktail for hybrid basalt-glass fiber. Test results indicated that the fracture energy of mix is highly dependent on both fiber dosage and temperature. Moreover, the hybrid fiber reinforced mixes showed the highest fracture energies in comparison with monotype fiber reinforced specimens with 1% fiber volume fraction. In general, hybridization has played a leading role in the improvement of mechanical properties and fracture behavior of mixes, while compared to monotype fiber reinforced specimens, hybridization has led to lower amounts of compressive strength.

• Preventive Nutrition and Food Science
• /
• v.5 no.4
• /
• pp.197-199
• /
• 2000

Clarified fruit and vegetable juices (apple, carrot and tangerine) were produced using ultrafiltration and their color change due to the mixing ratio were evaluated. clarification was carried out by passing he supernatant of extracted juice through a filter and also by using a membrane of molecular weight cut-off 10,000 Daltons to obtain the juice ultrafiltrates. The mixing ratio between apple and carrot juices was kept constant at 1:1 while increasing the amount of tangerine juice according to 10, 20, 30, 40 and 50% an stored at 4$^{\circ}C$ prior to the color measurement. Hue angle ({TEX}$h_{ab}${/TEX}) and {TEX}$L^{*}${/TEX}-value increased as the tangerine mixing ratio increased. The color difference indicated by ΔE-value also increased as the amount of tangerine increased indicating that the color of the mixed juice became pale and the changes were slight but distinctive. On the other hand, chroma ({TEX}$C^{*}${/TEX}), {TEX}$La{*}${/TEX}- and {TEX}$b^{*}${/TEX}-values decreased as the tangerine mixing ratio increased indicating that the color of the mixed juice became slightly more grayish and the samples were becoming les yellow. A simple mathematical model to predict each color characteristic is proposed.