• 한국시뮬레이션학회:학술대회논문집
• /
• 한국시뮬레이션학회 2001년도 The Seoul International Simulation Conference
• /
• pp.313-317
• /
• 2001

The approach to predict time series without neglecting the fluctuation in a short period is tried by using a digital FIR filter and a neural network. The differential waveform of the Nikkei average closing price is filtered by the FIR band-pass filter of 101 length. It is filtered into the five frequency bands of 0-1Hz, 1-2Hz, 2-3Hz, 3-4Hz and 4-5Hz by setting the sampling frequency 10Hz. The each filtered waveform is learned and forecasted by the neural network. The neural network of the back propagation method is adopted in the learning the waveform. By inputting the data of 20 days in the past, the prediction of 10 days ahead is carried out. After learning the time series of each frequency band by the neural network, the predicted data far each frequency band are obtained. The predicted waveforms of each frequency band are synthesized to obtain a final forecast. The waveform can be forecasted well as a whole.

• 농업과학연구
• /
• 제12권2호
• /
• pp.341-348
• /
• 1985

This study aims at analysis hog cycles and seasonal price patterns, and at develop the procedure for price forecasting based on the relative price ratios by farmers. Seasonal price patterns have been a persistent feature of hog markets. Some month have historically high price and other months historically low price. Hog price tend to be high in Feb, May, June, Sept, winter (Nov. to Jan.) and tend to be low in the other months. There have been four price cycles for 12 years, 1972-1984, the length of the hog price cycle has varied from 24 month to 42 months, with the irregular frequency. The increasing period of the price cycle lasted 23 months and the decreasing period of the price cycle lasted 13 months. Tables 2, 3, 4 in this study show average hog price ratios and the number of times price fall, rose for one, two, and three months ahead of each calendar month.

• 대한전기학회:학술대회논문집
• /
• 대한전기학회 2000년도 하계학술대회 논문집 A
• /
• pp.301-303
• /
• 2000

This paper presents Kohonen neural network and wavelet transform analysis based technique for industrial peak load forecasting for the purpose of peak demand control. Firstly, one year of historical load data were sorted and clustered into several groups using Kohonen neural network and then wavelet transforms are adopted using the Biorthogonal mother wavelet in order to forecast the peak load of one hour ahead. The 5-level decomposition of the daily industrial load curve is implemented to consider the weather sensitive component of loads effectively. The wavelet coefficients associated with certain frequency and time localization is adjusted using the conventional multiple regression method and the components are reconstructed to predict the final loads through a six-scale synthesis technique.

• 한국농공학회지
• /
• 제41권6호
• /
• pp.25-32
• /
• 1999

The purpose of this study is to analyse the accuracy of weather forecasts of temperature, precipitation probability , and sky condition and to evaluate the applicability of weather forecasts for the estimation of potential evapotranspiration for irrigation scheduling. Five weather station s were selected to compare forecasted and measured climatcal data. The error between forecasted and measured temperature was calculated and discussed. The accuracy of temperature forecast using relative frequency of the error was calculated . The temperature forecasting showed considerably high accuracy. Average sunshine hours for forecasted sky conditions were calculated and showed reasonable quality. From the reliability graphs, the forecasting precipation probabililty was reliable. Potential evapotranspirations were calculated and compared using forecast and measured temperatures. The weather forecast is considered usable for irrigation scheculing.

• 대한안전경영과학회지
• /
• 제3권2호
• /
• pp.113-121
• /
• 2001

Today, facilities are composed of many complex components or parts. Because of this characteristics, the frequency of failures is decreasing, but the strength of failures is increasing; therefore, the failure analysis about many complex components or parts was needed. In the former research about Fault Tree Analysis, failure data of similar facilities have been used for forecasting about target system or components, but in case that the system or components for forecasting failure is new or qualitative and quantitative data are given simultaneously, there are many difficulty in using Fault Tree Analysis with this incorrect failure data. Therefore, this paper deal with the Fault Tree Analysis method which be applied with Fuzzy theory in above case. In case that , therefore, if there is no the correct failure data, it is represented a system or components as qualitative variable. subsequently, it converted to the quantitative value using fuzzy theory, and the values used as the value for failure forecast.

• 한국도로학회논문집
• /
• 제10권4호
• /
• pp.139-150
• /
• 2008

본 연구는 도로연장측면에서 본선에 비해 상대적으로 연결로에서 발생하는 사고빈도가 높고, 교통사고가 증가하는 추세인 고속도로 연결로의 교통사고 예측모형의 개발에 초점을 두었다. 연결로 유형별(직결, 준직결, 루프)로 통계적으로 유의한 사고인자를 선정하고, 사고율과의 관계가 비선형 임을 분석하여 변수를 변형(Variables Transformation)하여 All possible 방식으로 예측모형을 개발하고, 통계적 진단 및 검증을 거쳐 유의성을 확인하였으며 이에 기존 개발 모형에 비해 예측력이 더욱 우수한 결과를 보였다. 개발된 사고예측모형은 보다 비용면에서 효율적이고, 안전한 트럼펫형 IC 연결로의 설계와 연결로 교통사고 원인분석에 활용될 수 있을 것으로 기대된다.

• 대한토목학회논문집
• /
• 제38권5호
• /
• pp.655-669
• /
• 2018

본 연구에서는 비정상성 Bayesian 빈도해석모형을 토대로 외부 기상인자에 의한 시변성을 고려할 수 있는 계절강수량 예측모형을 구축한 후 산정된 결과를 입력 자료로 하여 직접적으로 일단위 이하의 극치강수량을 상세화시킬 수 있는 베타 모델(four parameter beta, 4PB)을 연계하여 한강 및 금강유역의 미래 계절 강수량 전망 및 일단위 이하의 확률강수량을 도출하였다. 모형의 적합성 검증을 위하여 2014~2017년의 모의된 사후 확률분포 값과 관측치를 비교하였다. 그 결과 계절강수량 모의에서 한강은 관측 값의 최대 약 86.3%, 금강은 약 98.9% 일치하는 것을 확인할 수 있었다. 지속시간별 극치강우량은 약 65.9~99.7%의 정확성을 나타냈다. 이에 본 연구에서 산정한 결과는 기상변동성을 다양한 시간규모에서 고려하기 위한 정보로 활용할 수 있을 것으로 판단된다.

• 한국지능시스템학회논문지
• /
• 제27권2호
• /
• pp.179-186
• /
• 2017

기술예측은 과거부터 현재까지의 기술개발 결과를 수집, 분석하여 특정 기술의 미래 추세 및 상태를 예측하는 것이다. 일반적으로 특허는 현재까지의 기술개발 결과를 가장 잘 가지고 있다. 왜냐하면 특허에 포함된 세부 기술은 일정기간 동안 배타적 권리가 법에 의해 보장되기 때문이다. 따라서 특허 데이터의 분석을 이용한 기술예측의 다양한 연구가 진행되었다. 특허문서의 분석을 위하여 널리 사용되는 특허 키워드 데이터는 주로 기술키워드에 대한 빈도 값으로 이루어진다. 기존의 많은 특허분석에서는 회귀분석, 박스-젠킨스 모형 등 연속형 데이터분석 기법이 적용하였다. 하지만 빈도 데이터는 이산형 데이터이기 때문에 이산형 데이터분석 방법을 사용해야 한다. 본 연구에서는 이와 같은 문제점을 해결하기 위하여 베이지안 포아송 이산모형을 이용한 특허분석 방법을 제안한다. 연구방법의 성능평가를 위하여 지금까지 출원, 등록된 애플의 전체특허를 분석하여 향후 기술을 예측하는 사례분석을 수행한다.

• 한국수자원학회논문집
• /
• 제41권10호
• /
• pp.1023-1034
• /
• 2008

최근 몇 년간 전 세계에 걸쳐 폭풍우와 관련한 자연재해는 그 규모와 빈도에 있어서 상당히 증가하고 있는 추세다. 특히, 우리나라는 강수의 대부분이 여름철에 집중되어 있어 이러한 태풍, 폭우 그리고 국지성 집중호우 등과 같은 자연재해로 인한 피해가 더욱 심각하다. 이러한 현상은 대기 중 이산화탄소 농도의 증가로 인한 지구온난화와 엘리뇨 등으로 인하여 앞으로도 더욱 빈번해질 것으로 전망된다. 따라서 이와 같은 폭풍우로 인한 피해를 줄이기 위하여 본 연구에서는 기상레이더를 이용한 단시간 강우예측 모형을 개발하였다. 본 연구는 3차원으로 생산되는 레이더 자료를 2차원 CAPPI(Constant Altitude Plan Position Indicator)로 변환, 강우의 이동방향과 이동속도 예측, 현업보정을 이용한 2차원 강우량 산정으로 구성되어 있다. 연구결과 기상레이더를 이용한 국지성 호우의 단시간 강우예측 가능성을 제시하였으며 향후 홍수 예 경보시스템과 연계한다면 홍수 관리 및 피해 경감에 기여할 것으로 판단된다.

• 한국수자원학회:학술대회논문집
• /
• 한국수자원학회 2002년도 학술발표회 논문집(I)
• /
• pp.43-50
• /
• 2002

Accurate quantitative forecasting of rainfall for basins with a short response time is essential to predict streamflow and flash floods. Previously, neural networks were used to develop a Quantitative Precipitation Forecasting (QPF) model that highly improved forecasting skill at specific locations in Pennsylvania, using both Numerical Weather Prediction (NWP) output and rainfall and radiosonde data. The objective of this study was to improve an existing artificial neural network model and incorporate the evolving structure and frequency of intense weather systems in the mid-Atlantic region of the United States for improved flood forecasting. Besides using radiosonde and rainfall data, the model also used the satellite-derived characteristics of storm systems such as tropical cyclones, mesoscale convective complex systems and convective cloud clusters as input. The convective classification and tracking system (CCATS) was used to identify and quantify storm properties such as life time, area, eccentricity, and track. As in standard expert prediction systems, the fundamental structure of the neural network model was learned from the hydroclimatology of the relationships between weather system, rainfall production and streamflow response in the study area. The new Quantitative Flood Forecasting (QFF) model was applied to predict streamflow peaks with lead-times of 18 and 24 hours over a five year period in 4 watersheds on the leeward side of the Appalachian mountains in the mid-Atlantic region. Threat scores consistently above .6 and close to 0.8 ∼ 0.9 were obtained fur 18 hour lead-time forecasts, and skill scores of at least 4％ and up to 6％ were attained for the 24 hour lead-time forecasts. This work demonstrates that multisensor data cast into an expert information system such as neural networks, if built upon scientific understanding of regional hydrometeorology, can lead to significant gains in the forecast skill of extreme rainfall and associated floods. In particular, this study validates our hypothesis that accurate and extended flood forecast lead-times can be attained by taking into consideration the synoptic evolution of atmospheric conditions extracted from the analysis of large-area remotely sensed imagery While physically-based numerical weather prediction and river routing models cannot accurately depict complex natural non-linear processes, and thus have difficulty in simulating extreme events such as heavy rainfall and floods, data-driven approaches should be viewed as a strong alternative in operational hydrology. This is especially more pertinent at a time when the diversity of sensors in satellites and ground-based operational weather monitoring systems provide large volumes of data on a real-time basis.