Probabilistic explosion risk analysis for offshore topside process area. Part II: Development of gas cloud multivariate frequency distribution (MVFD)

Jin, YanLin,Jang, Beom-Seon Elsevier 2018 Journal of loss prevention in the process industri Vol.51 No.-

<P><B>Abstract</B></P> <P>Flammable gas cloud frequency distribution forms a key part of the CFD-based probabilistic explosion risk analysis (ERA). As part of this series work, a new type of gas cloud frequency distribution, namely, ignited gas cloud frequency distribution has been proposed in the previous study (Part I). On that basis, an advanced gas cloud frequency distribution is to be introduced continuously, which is the ultimate goal of this series work. The current study contributes to proposing a multivariate frequency distribution (MVFD), which can provide both gas cloud size and position to investigate explosion scenarios. As a matter of fact, the consequences of explosion accidents may vary greatly, depending on variables such as ignition position, gas cloud size, position and shape etc. So far, except for the gas cloud size that can be provided by the gas cloud frequency distribution, the remaining variables are more likely to be determined by engineering judgement and experience. Though this process follows standard guidelines or recommended practices, the determined variables can vary widely depending on engineers. In the present study, however, the gas cloud position is determined probabilistically as well. In order to input the gas cloud position into the proposed MVFD, the volumetric center of the equivalent gas cloud (ESC) is used to quantify the cloud position. To this end, the methodology introduced in Part I is consistently adopted in this paper, but the volumetric center is additionally monitored in the whole transient process of cloud propagation. Using the proposed distribution, the explosion scenarios can be investigated without the subjective engineering judgment of the gas cloud position, and therefore one can expect that the investigated explosion scenarios become more accurate than before. This paper provides a detailed procedure on how to obtain the proposed distribution and then presents case studies to demonstrate how the ERA results can change with the choice of the gas cloud position. Through the case studies, it is proven that the proposed MVFD is more reasonable for evaluating correct explosion design accidental loads.</P> <P><B>Highlights</B></P> <P> <UL> <LI> A new type of gas cloud frequency distribution is proposed to consider both the position and size of gas clouds in a probabilistic manner. </LI> <LI> The position of a gas cloud is defined as the volumetric center of a corresponding equivalent stoichiometric gas cloud. </LI> <LI> The entire footprints of gas cloud propagation are reflected into the proposed distribution. </LI> <LI> Investigated explosion scenarios are consistent with the results of dispersion simulations, and become more realistic than before. </LI> </UL> </P>

혼합분포형을 이용한 한국의 연최대 풍속 분석

신주영,안현준,정창삼,안정환 한국방재학회 2018 한국방재학회논문집 Vol.18 No.7

Recently, many studies on wind speed frequency analysis reported that mixture distribution is appropriate for the distribution of wind speed data. Since the annual maximum wind speed data are a subsample of wind speed data, the characteristics of the annual maximum wind speed data are highly likely related to those of wind speed data. Hence, in this study, the fit of mixture distribution for annual maximum wind speed data measured in Korea was assessed. A two-component Gumbel mixture distribution was employed, and an expectation-maximization algorithm was used as fitting method. The fit of the mixture distribution was compared with those of the Gumbel, Weibull, Generalized Extreme Value, and Kappa distributions, based on several goodness-of-fit measurements. The results show that the fit of the mixture distribution for the annual maximum wind speed in Korea was superior to the fits of other distributions. Additionally, the mixture distribution led to good fits also at those stations where it was not selected as the most appropriate distribution. 최근 풍속자료의 빈도해석 연구에서 풍속자료의 분포형으로 혼합확률분포이 적합하다는 결과가 보고되고 있다. 연 최대 풍속자료는 풍속자료 중에서 연 최대 값을 뽑아 낸 자료로, 풍속자료의 특성에 따라 연 최대 풍속자료의 특성이 결정될 가능성이 높다. 그래서, 본 연구에서는 국내 연 최대 풍속자료에 대한 혼합분포형의 적합성을 평가였다. 혼합분포형으로는 Gumbel 분포형을 두 개를 합성한 분포형을 사용하였고, 혼합분포형의 매개변수를 추정하기 위해서, expectation-maximization 알고리즘을 적용하였다. 추정된 혼합분포형은 다양한 적합도 평가 기준을 토대로 Gumbel, Weibul, Generalized Extreme Value, Kappa 분포형과 국내 풍속자료에 대한 적합도를 비교하였다. 평가결과, 혼합분포형이 극치 분석에 사용되는 분포형들보다 국내 연 최대 풍속자료에 적합한 것으로 나타났고, 혼합분포모형이 최적분포형으로 선정되지 않은 지점에서도 높은 적합도를 보이는 것으로 확인하였다.

엔트로피 모형을 활용한 고속철도 역세권 통행분포 추정에 관한 연구

조항웅(Cho, Hangung),김시곤(Kim, Sigon),김진환(Kim, Jinhowan),전상민(Jeon, Sangmin) 대한토목학회 2012 대한토목학회논문집 D Vol.32 No.6D

지난 2004년 4월 경부고속철도 1단계 개통이후, 2단계 사업은 2010년 11월 개통하였으며, 고속철도 개통이후 타 교통수단에 비해 속도 경쟁의 장점을 가지고 있어 고속철도의 수요는 계속 증가하고 있다. 이러한 고속철도의 개통은 인구의 이동, 기업의 입지, 공간구조의 개편 등과 같은 사회적, 경제적, 교통적인 변화를 주도하고 있는 실정이며, 특히 고속철도의 고속 운행으로 지역간의 이동시간을 단축한다는 점에서 고속철도 수요는 계속적인 증가 추세로 전망된다. 본 연구에서는 고속철도 서울역 설문조사의 데이터를 이용한 EMME/2 프로그램의 2-Dimentional Blancing을 활용한 고속철도 역 접근수단별 통행 분포 모형의 파라메타 추정을 통하여 조사 통행분포를 추정 통행분포와 같이 재현하고자 하였으며, 분석 결과 접근수단별로 파라메타(?)는 승용차 0.0395, 버스 0.0390, 지하철 0.0415, 택시 0.0650으로 분석되었고, 통행거리빈도분포(Trip Length Frequency Distribution: TLFD)를 기준으로 조사치와 모형치를 비교한 결과 R²는 승용차 0.909, 버스0.923, 지하철 0.922, 택시 0.745로 조사치와 모형치는 유사한 것으로 분석 되었으며, F검증 결과 P값이 모두 0.05보다 매우 작게 분석되어 95%신뢰수준으로 유의할 만 한 것으로 판단되었다. 통행거리빈도분포를 5㎞ 단위로 설정하여 분석 하였으나, 향후에는 통행거리빈도분포를 중죤단위에서 소죤단위(행정동)로 세분화 연구가 필요하며, 통행거리 0~5㎞ 구간의 분포을 반영할수 있는 결합함수(Combined function}을 활용한 중력모형과 3-Dimentional Blancing을 적용한 연구가 필요 할 것으로 판단된다. KTX step 1 April 2004, after the opening, the second phase of the project was opened in November 2010. High-speed rail after the opening and continue to increase the demand of high-speed rail, Have the speed of competitive advantage compared too the means of transportation. The opening of these high-speed rail has led to changes of the move, the company’s position, and the spatial structure of the population of reorganization, such as the social, economic, transportation. In this study, survey data using the High Speed Rail Station EMME/2 of the program to take advantage of the 2-Dimentional Blancing trip distribution to investigate the passage through the trip distribution by the estimation of the parameters of the model to estimate the distribution of the means of access and high-speed rail station to reproduce and Analysis of the results by means of access parameters (?) autos 0.0395, buses 0.0390, subway 0.0650, taxi 0.0415, the frequency distribution (Trip Length Frequency Distribution: TLFD) were analyzed survey data value model with the results of comparing R² cars analysis and model values similar survey data 0.909 bus 0.923, subway 0.745 to 0.922, taxi, F test P value analysis is smaller than 0.05 at the 95% confidence level as a note that was judged to have been. Trip frequency distribution analysis, but in the future, set the unit to 5㎞-trip frequency distribution middle zone units from small zone units (administrative district) segmentation research is needed, and can reflect the trip distance 0~5 ㎞ interval combined function to take advantage of the gravity model and the 3-Dimentional Blancing applied research is needed to be considered.

Bayesian 기법을 이용한 혼합 Gumbel 분포 매개변수 추정 및 강우빈도해석 기법 개발

최홍근(Choi, Hong-Geun),오랑치맥솜야(Uranchimeg, Sumiya),김용탁(Kim, Yong-Tak),권현한(Kwon, Hyun-Han) 대한토목학회 2018 대한토목학회논문집 Vol.38 No.2

우리나라의 기후 지형적 특성에 따라 연강수량의 50% 이상이 여름철에 내린다. 이러한 짧은 기간에 집중적으로 내리는 강수량 조건하에 수공구조물을 설계할 경우 대부분 극치빈도분석을 활용한다. 특히 우리나라의 경우 Gumbel 분포를 활용한 극치빈도분석을 많이 이용한다. 하지만, 최근 이상기후로 인하여 전세계적으로 강수량의 특징이 급격히 변하고 있으며, 우리나라 연강수량 특징도 바뀌고 있다. 즉, 기존의 단일 분포형으로 재현이 가능했던 수문기상 자료들이 혼합분포형의 특징을 가지게 되었으며 이러한 변화를 고려할 수 있는 극치빈도분석 개발이 요구되고 있는 실정이다. 본 연구에서는 두 개 이상의 첨두를 가지는 형태의 극치강수량 자료에 대해서 기존의 단일 Gumbel 분포형 기반 극치빈도분석과 혼합 Gumbel 분포형 기반의 극치빈도분석 결과를 비교하였다. 확률분포의 매개변수 산정시 우도함수를 Bayesian 기법을 통해 산정하여 각 분포형의 Bayesian information criterion (BIC) 값을 비교하였다. 분석한 결과, 앞서 제안된 혼합 Gumbel 분포형은 하나의 첨두를 가지는 단일 Gumbel 분포형에서 반영되지 못한 꼬리(tail)부분의 이중첨두 부분의 거동을 효과적으로 모의하는 것을 확인할 수 있었다. 결과적으로 설계강수량을 추정할 때 보다 신뢰성있는 접근이 가능하였다. 이러한 점에서 우리나라 극치강우자료 분석시 기존 단일분포기반의 빈도해석기법에 대안으로 적용이 가능할 것으로 판단된다. More than half of annual rainfall occurs in summer season in Korea due to its climate condition and geographical location. A frequency analysis is mostly adopted for designing hydraulic structure under the such concentrated rainfall condition. Among the various distributions, univariate Gumbel distribution has been routinely used for rainfall frequency analysis in Korea. However, the distributional changes in extreme rainfall have been globally observed including Korea. More specifically, the univariate Gumbel distribution based rainfall frequency analysis is often fail to describe multimodal behaviors which are mainly influenced by distinct climate conditions during the wet season. In this context, we purposed a Gumbel mixture distribution based rainfall frequency analysis with a Bayesian framework, and further the results were compared to that of the univariate. It was found that the proposed model showed better performance in describing underlying distributions, leading to the lower Bayesian information criterion (BIC) values. The mixed Gumbel distribution was more robust for describing the upper tail of the distribution which playes a crucial role in estimating more reliable estimates of design rainfall uncertainty occurred by peak of upper tail than single Gumbel distribution. Therefore, it can be concluded that the mixed Gumbel distribution is more compatible for extreme frequency analysis rainfall data with two or more peaks on its distribution.

한국어와 영어 기본 단어들의 빈도에 대한 계량적 분석

김선회 ( Kim¸ Sun-hoi ) 한국현대언어학회 2021 언어연구 Vol.37 No.2

The Journal of Studies in Language 37.2, 127-148. This study focuses on the frequencies of basic words in Korean and English. Their distributions were analyzed quantitatively to compare Korean and English. The correlation analysis shows that the inter-correlation between Korean and English is very high. It implies that the linguistic-cultural difference between these two languages rarely affects the use of basic words. The analysis of visualized frequency distributions shows that the frequencies of basic words exhibit heavy tailed distributions and seem to follow the power-law distribution in both languages. However, the quantitative analysis tells us that their frequency distributions do not follow the power-law distribution but the log-normal distribution. The result suggests that selecting probability distribution best fitted to frequency data should depend on well-established statistical procedures, but not only on visually inspecting of plotted distributions. (Chung-Ang University)

학교수학에서 통계적 분포에 대한 지도 분석 연구: 도수분포 그래프를 중심으로

고은성,탁병주 한국초등수학교육학회 2023 한국초등수학교육학회지 Vol.27 No.4

This study analyzed how distributions are taught in the statistics unit of the Korean mathematics curriculum, focusing on bar graphs in elementary schools and stem-and-leaf plots, histograms, and frequency polygons in middle schools. Questions in elementary and middle school mathematics textbooks for teaching these graphs were analyzed from the perspective of reasoning about distribution and categorized into types (local view, global view, and other) to identify their characteristics. As a result of the analysis, we identified eight types of questions that encourage a global view (representing on a graph, considering categories, noting a specific range, listing in order of frequency, finding a graph shape, describing features of a distribution, comparing two groups, and making graph-based decisions). Second, in bar graph unit in the elementary school, the proportion of questions in the order of "other", "local view", and "global view" was high. However, the proportion of questions in the "other" category varied significantly between textbooks, and the proportion of questions in the "local view" category varied significantly between textbooks. On the other hand, in middle school, the stem-and-leaf plot, histogram, and frequency polygon units had a high percentage of questions categorized as "global view," while the percentage of questions categorized as "local view". Based on the results of this study, we propose some implications for improving mathematics curriculum and textbooks for teaching of distribution.

출판유통 RFID 적용에 관한 연구

안종윤,강경식 대한안전경영과학회 2004 대한안전경영과학회지 Vol.6 No.1

Greeting the globalization and information period of 21C, the improvement of the distributional system structure and modernization of publishing distribution are the seriously required periodic assignments for the publishing industry that is a national knowledge and culture industry to be equipped with the core competitiveness. First, it is to found a large distributional organization that is able to provide total logistics services in time that is essential to the publishing company and bookstore. Second, it is to have a logistics service to act up to the variable requirements of customers such as publishing companies and bookstores under the rapidly changing distribution environment. Third, it is to provide the standardized information of publishing distribution. It is needed to establish a total distributional center in collaboration to satisfy the requirements of publishing industry as above. By doing this work, structure of the publishing distribution system can be improved and its constitution can be strengthen, and it is possible not only to construct a stable supply system but also to realize an information-based system in the publishing industry with the construction of information system and its utilization. Furthermore, the application of RFID in the publishing distribution is essentially required to provide the symbolic model and logistics infrastructure into the 'Paju publishing culture information industrial complex'. It could be possible by growing the international recognition regarding distribution environment and strengthening the international competitiveness in the period of the ""Open door".

Distribution of the Mandarin vowels in typological perspective

Ok Joo Lee,Yan Xiong 경희대학교 언어정보연구소 2021 언어연구 Vol.38 No.2

The present study is designed to examine the distribution of the Mandarin vowels from a typological perspective. Despite little research on this topic, earlier literature has reported that Mandarin is an exceptional language which contains a larger number of uncommon marked vowels (e.g., Crothers 1978). The purpose of the present study is twofold. The primary aim is to achieve a better understanding of the typological features of the Mandarin vowel system by analyzing the frequency distribution of the Mandarin vowels and the relation between markedness and distribution. Another goal is to examine the vowel distribution differences between two types of language data, namely, lexicon and natural speech. Results of quantitative analyses reveal that with the most unmarked peripheral vowels /a/, /i/, and /u/ being the most frequently occurring ones, the overall vowel distribution patterns in Mandarin lexicon conform to the cross-linguistic tendencies of vowel distribution and markedness. While the tendency of most vowels favoring Tone 4 is found, the relative percentage of Tone 1 and Tone 3 occurrences appears to be higher in the low vowels than the high and mid vowels in lexicon. However, different distribution patterns emerge when comparing lexicon and natural speech, in that the mid vowels (i.e., /ə/, /ɤ/, /o/) and apical vowel (i.e., /ʅ/) appear at significantly higher frequencies in natural speech. The differences can be explained by the fact that these vowels occur in some high frequency words (e.g., ‘de’, ‘wǒ’, ‘shì’). Findings of this study suggest that different claims in previous studies of phonological typology are in part attributed to the nature of the language data examined (e.g., phonemic inventory, lexicon, natural speech).

Praat과 R로 분석한 한국인 대화 음성 말뭉치의 fundamental frequency(f0)값 분포

양병곤 한국음성학회 2023 말소리와 음성과학 Vol.15 No.3

This study examines the fundamental frequency(f0) distribution of 2,740 Korean speakers in a dialogue speech corpus. Praat and R were used for the collection and analysis of acoustical f0 data after removing extreme values considering the interquartile f0 range of the intonational phrases produced by each individual speaker. Results showed that the average f0 value of all speakers was 185 Hz and the median value was 187 Hz. The f0 data showed a positively skewed distribution of 0.11, and the kurtosis was –0.09, which is close to the normal distribution. The pitch values of daily conversations varied in the range of 238 Hz. Further examination of the male and female groups showed distinct median f0 values: 114 Hz for males and 199 Hz for females. A t-test between the two groups yielded a significant difference. The skewness representing the distribution shape was 1.24 for the male group and 0.58 for the female group. The kurtosis was 5.21 and 3.88 for the male and female groups, and the male group values appeared leptokurtic. A regression analysis between the median f0 and age yielded a slope of 0.15 for the male group and –0.586 for the female group, which indicated a divergent relationship. In conclusion, a normative f0 distribution of different Korean age and sex groups can be examined in the conversational speech corpus recorded by a massive number of participants. However, more rigorous data might be required to define a relation between age and f0 values.

최소자승법을 이용한 원판형 절리의 직경분포와 체적빈도 추정에 관한 연구

송재준(Jae-Joon Song) 한국암반공학회 2005 터널과지하공간 Vol.15 No.2

이 연구에서는 최소자승법을 이용하여 절리의 직경분포를 추정하는 방법을 개발하였다. 이전에 Song and Lee가 제안한 방법에서는 현장에서 조사한 양끝내포선(contained trace)분포로부터 무한 조사창에서 정의되는 절리선(joint trace) 길이 분포를 먼저 구하고 이 후에 직경분포를 구하게 된다 그러나 새로 제안한 방법을 사용하면 중간 추정과정없이 현장에서 얻은 양끝내포선 분포로부터 바로 절리의 직경분포를 구할 수 있다 이전의 방법과 비교할 때 새로 제안된 방법은 표본의 크기가 작을 때 조금 더 높은 추정정밀도를 보이며, 직경분포를 추정하는 과정에서 절리의 기하학적 파라미터의 하나인 체적빈도(volumetric frequency)도 제공한다. 새로운 추정법의 검증을 위해 Monte Carlo 시뮬레이션을 적용하였다. An estimation technique of the joint diameter distribution using the least square method is suggested. When utilizing the technique by Song and Lee, the diameter distribution would be obtained only from the trace length distribution defined in an infinite window after the trace length distribution is estimated from the contained trace length distribution. With the new technique, however, the diameter distribution can be directly obtained from the sample histogram of the contained trace lengths. Compared with the previous technique, it shows a more accurate result for small sizes of joint samples and provides the joint geometry parameter of volumetric frequency. Verification of this new technique was completed by using Monte Carlo simulations.