• Journal of Korean Society of Transportation
• /
• v.23 no.5 s.83
• /
• pp.147-155
• /
• 2005

The purpose of this paper is to develop a method for estimation of reliable path-travel time using data obtained from the toll collection system on freeways. The toll collection system records departure and arrival time stamps as well as the identification numbers of arrival and destination tollgates for all the individual vehicles traveling between tollgates on freeways. Two major issues reduce accuracy when estimating path-travel time between an origin and destination tollgate using transaction data collected by the toll collection system. First, travel time calculated by subtracting departure time from arrival time does not explain path-travel time from origin tollgate to destination tollgate when a variety of available paths exist between tollgates. Second, travel time may include extra time spent in service and/or rest areas. Moreover. ramp driving time is included because tollgates are installed before on-ramps and after off-ramps. This paper describes an algorithm that searches for arrival time when departure time is given between tollgates by a Progressive Iterative Forward and Backward (PIFAB) search method. The algorithm eventually produces actual path-travel times that exclude any time spent in service and/or rest areas as well as ramp driving time based on a link-based procedure.

• Journal of the Korean Society of Marine Environment & Safety
• /
• v.25 no.6
• /
• pp.809-818
• /
• 2019

Since January 1, 2020, the International Maritime Organization (IMO) has put in place strong regulations to reduce air pollution caused by ships by lowing the upper limit of ship fuel oil sulfur content from 3.5% to 0.5% for ships passing through all sea areas around the world. Although it is important to reduce air pollutants by using fuel oil with low sulfur content, reducing the amount of energy waste through the economic operation of a ship can also help reduce air pollutants. Ships can follow designated routes accurately even under the influence of noise using autopilot systems. However, regardless of their quality, the performance of these systems is af ected by noise; heading angles with added measurement noise from the gyroscope are input into the autopilot system and degrade its performance. A technique to solve these problems reduces noise effects through the application of a Kalman filter, which is widely used in condition estimation. This method, however, cannot completely eliminate the effects of noise. Therefore, to further improve noise removal performances, in this study we propose a better denoising method than the Kalman filter technique by applying a multi-layer perceptron (MLP) in forward direction motion and a Kalman Filter in rotation motion. Simulations show that the proposed method improves forward direction motion by preventing the malfunction of a rudder more so than merely using a Kalman Filter.

• Journal of Fisheries and Marine Sciences Education
• /
• v.9 no.2
• /
• pp.179-187
• /
• 1997

Doppler log and current meter are based on the measurment of the doppler effect. Ship's speed over the ground measured by means of doppler log effect of underwater ultrasonic intends to get infulence by the difference of sound velocity, the ship's course, the sea bottom inclination, the trim and tranducer installation etc. This paper investigated on the error of speed over the ground by change of the trim comparing the real speed obtained by the mile post with the speed of single beam type doppler log on the trainning ship Kyeongyangho. The results are as follows ; 1. Indicating speed of doppler log is very much greatly influenced on variation of trim, but the real speed is less affected variation by trim. 2. The range of variation and error of speed over ground are smallest when ship's trim is 2.15m trim by the stern.

• Journal of the Mineralogical Society of Korea
• /
• v.20 no.1 s.51
• /
• pp.35-46
• /
• 2007

Shinyemi skarn deposits occur as Fe-Mo skarn type and Pb-Zn-Cu hydrothermal replacement type along the contact between Cretaceous Shinyemi granitoids and Cambro-Ordovician mixed limestone and dolostone sequence of the Choseon Supergroup. In the lower part of Western Shinyemi ore body two stages of skarn formation have been observed: the early, stage I (magnesian) skarn with Fe mineralization and the late, stage II(calcic) skarn with Mo mineralization. The stage I skarn spatially is overprinted by stage II skarn. The stage I skarn is predominantly composed of olivine, magnetite and diopside whereas, the stage II skarn is dominated by hedenbergite and garnet. The skarnification process occurred in two stages, both prograde and retrograde for stage I and stage II skarns. In stage I, the prograde skarns, mainly composed of anhydrous silicate minerals, were formed at relatively higher temperatures (about $400\;to\;550^{\circ}C$) under low $CO_{2}$ fugacity ($X_{CO2}<0.1$) conditions. On the other hand, the retrograde skarns that consisted of hydrous minerals were formed at lower temperatures (about $300\;to\;400^{\circ}C$).

• Journal of Animal Science and Technology
• /
• v.47 no.4
• /
• pp.491-500
• /
• 2005

To test applicability to the Hanwoo traceability system, twenty microsatellite markers were selected and analyzed. MSA, CERVUS, FSTAT, GENEPOP, API_CALC and PHYLIP software was employed serially to estimate heterozygosity, polymorphic information content, F-statistics, identity probability, exclusion probability and genetic distance. Eleven microsatellite markers(TGLA53, TGLA227, ETH185, TGLA122, BM4305, INRA23, ILSTS013, BMS1747, BM2113, BL1009, and ETH3) were selected based on their high heterozygosity values. Identity probability using these markers is one hundred times higher than when using StockMakersTM of Applied Biosystems. This indicates the selected microsatellite markers are appropriate and effective for use in the Hanwoo traceability system. Additionally, estimates of DA genetic distance and pairwise-FST can be utilized to identify genetic relationships between adjacent farms.

• Journal of agriculture & life science
• /
• v.44 no.2
• /
• pp.17-28
• /
• 2010

This study looked into how much maternal genetic effect influenced on economic traits through estimation of genetic parameter and heritability over swine's economic traits by maternal animal model using GGP farm examination data of total 31,455 swine of Duroc species, Landrace species and Yorkshire species that were born between 2000 and 2008. As a result of significance test over each factor in surveyed all traits, high significance was approved in the effect of breed, gender, the date of swine's birth, the season swine's born, and difference in delivery in every trait (p<0.01). It is considered that it would be possible to get more efficient improvement effect provided correlation between additive genetic effect and maternal genetic effect as well as maternal genetic effect according to breed, traits, and improvement direction are properly considered as negative covariance existed between additive genetic distribution and maternal genetic distribution presumed for traits by each breed and their genetic relation also showed mostly strong negative correlation.

• Journal of Animal Science and Technology
• /
• v.53 no.1
• /
• pp.29-34
• /
• 2011

Using the materials collected from nine farms in a three-way cross system to produce commercial pigs produced from F1 sows (Landrace $\times$ Large White) $\times$ Duroc, the power of individual discrimination and parentage of the 13 microsatellite (MS) marker set that has been suggested for individual/brand identification (traceability) was empirically tested. Initially, genotypes of the parental population ($F_1$ sows and Duroc), and commercial pigs were determined and the genotype frequency and polymorphic index were estimated using the Cervus 2.0 program. The probability of identity among genotypes of random individuals, that random half sibs and that of full sib individuals, based on the genotypes from 91 $F_1$ sows and Duroc were expected to be $4.94{\times}10^{-34}$, $8.16{\times}10^{-23}$ and $2.01{\times}10^{-08}$, respectively, using the API-CALC version 1.0 program. When commercial pigs were included, the estimates increased to $3.74{\times}10^{-35}$, $5.48{\times}10^{-25}$ and $2.96{\times}10^{-11}$, respectively. For the empirical verification of the estimated powers of individual discrimination and parentage, the parentage test was performed for 452 commercial pigs using PAPA version 2.0, and individuals with the same genotype were investigated using the Cervus version 2.0 program. Parents for all commercial pigs were successfully estimated and no identical individual was identified in the pedigree. Although the individual discriminating power was not fully verified because of the lack of individuals corresponding with the theoretical power, the 100% efficiency of parentage test was clearly confirmed. Therefore, we believe that the 13 MS marker set in conjunction with management record/information for the pig production kept in a farm/brand should be useful in the pork traceability in a brand unit.

• Journal of agriculture & life science
• /
• v.44 no.5
• /
• pp.55-64
• /
• 2010

This study was conducted to estimate genetic parameters for reproductive traits using multivariate animal models in Yorkshire breed. For the study, 4,989 records for litter traits collected between the year 2001 and 2005 from Yorkshire pigs in K GGP were used. The effects of environmental factors such as farrowing year, parity, weaning to estrus interval (WEI), and suckling period were statistically significant (p<0.05), but farrowing season was not significant, for reproductive traits. The estimates genetic correlations and phenotypic correlations in total number of born and number of suckling, was shown to highly correlated. The genetic correlations were higher than phenotypic correlation. The estimates of heritabilities for reproductive traits, considering permanent environment effects (PE) were much lower than those obtained when permanent environment effects were not considered (NPE) in the model. The estimates of heritabilities were 0.240 and 0.076 for total number of born and 0.187 and 0.096 for number of suckling in NPE, and PE, respectively. These results itivcate that PE should be considered in the statistical mode to estimate more acco ate breeding values.

• The Journal of the Korea Contents Association
• /
• v.6 no.1
• /
• pp.23-30
• /
• 2006

A clustering method for temporal data takes a model-based approach. This uses automata based model for each cluster. It is necessary to construct global models for a set of data in order to elicit individual models for the cluster. The preparation for building individual models is completed by determining the number of clusters inherent in the data set. In this paper, BIC(Bayesian Information Criterion) approximation is used to determine the number clusters and confirmed its applicability. A search technique to improve efficiency is also suggested by analyzing the relationship between data size and BIC values. A number of experiments have been performed to check its validity using artificially generated data sets. BIC approximation measure has been confirmed that it suggests best number of clusters through experiments provided that the number of data is relatively large.

• The Journal of the Korea Contents Association
• /
• v.7 no.6
• /
• pp.22-30
• /
• 2007

Most real word systems such as world economy, stock market, and medical applications, contain a series of dynamic and complex phenomena. One of common methods to understand these systems is to build a model and analyze the behavior of the system. In this paper, we investigated methods for best clustering over time series data. As a first step for clustering, BIC (Bayesian Information Criterion) approximation is used to determine the number of clusters. A search technique to improve clustering efficiency is also suggested by analyzing the relationship between data size and BIC values. For clustering, two methods, model-based and similarity based methods, are analyzed and compared. A number of experiments have been performed to check its validity using real data(stock price). BIC approximation measure has been confirmed that it suggests best number of clusters through experiments provided that the number of data is relatively large. It is also confirmed that the model-based clustering produces more reliable clustering than similarity based ones.