• 한국경영과학회지
• /
• 제23권1호
• /
• pp.109-141
• /
• 1998

Many processes in the industrial realm exhibit sstochastic and nonlinear behavior. Consequently, an intelligent system must be able to nonlinear production processes as well as probabilistic phenomena. In order for a knowledge based system to control a manufacturing processes as well as probabilistic phenomena. In order for a knowledge based system to control manufacturing process, an important capability is that of prediction : forecasting the future trajectory of a process as well as the consequences of the control action. This paper examines the robustness of data mining tools under varying levels of noise while predicting nonlinear processes, includinb chaotic behavior. The evaluated models include the perceptron neural network using backpropagation (BPN), the recurrent neural network (RNN) and case based reasoning (CBR). The concepts are crystallized through a case study in predicting a chaotic process in the presence of various patterns of noise.

• 대한안전경영과학회지
• /
• 제11권1호
• /
• pp.205-211
• /
• 2009

In reality, distribution planning for a supply chain is established using a certain probabilistic distribution estimated by forecasting. However, in general, the demands used for an actual distribution planning are of deterministic value, a single value for each of periods. Because of this reason the final result of a planning has to be a single value for each period. Unfortunately, it is very difficult to estimate a single value due to the inherent uncertainty in the probabilistic distribution of customer demand. The issue addressed in this paper is the selection of single demand value among of the distributed demand estimations for a period to be used in the distribution planning. This paper proposes an efficient demand selection scheme for minimizing total inventory costs while satisfying target service level under the various experimental conditions.

• KIEE International Transactions on Power Engineering
• /
• 제2A권3호
• /
• pp.95-101
• /
• 2002

This paper illustrates a new fuzzy effective load model for probabilistic and fuzzy production cost simulation of the load point of the composite power system. A model for reliability evaluation of a transmission system using the fuzzy set theory is proposed for considering the flexibility or ambiguity of capacity limitation and overload of transmission lines, which are subjective matter characteristics. A conventional probabilistic approach was also used to model the uncertainties related to the objective matters for forced outage rates of generators and transmission lines in the new model. The methodology is formulated in order to consider the flexibility or ambiguity of load forecasting as well as capacity limitation and overload of transmission lines. It is expected that the Fuzzy CMELDC (CoMposite power system Effective Load Duration Curve) proposed in this study will provide some solutions to many problems based on nodal and decentralized operation and control of an electric power systems in a competitive environment in the future. The characteristics of this new model are illustrated by some case studies of a very simple test system.

• 한국경영과학회:학술대회논문집
• /
• 한국경영과학회 1997년도 추계학술대회발표논문집; 홍익대학교, 서울; 1 Nov. 1997
• /
• pp.181-184
• /
• 1997

Practical applications of learning systems usually involve complex domains exhibiting nonlinear behavior and dilution by noise. Consequently, an intelligent system must be able to adapt to nonlinear processes as well as probabilistic phenomena. An important class of application for a knowledge based systems in prediction: forecasting the future trajectory of a process as well as the consequences of any decision made by e system. This paper examines the robustness of data mining tools under varying levels of noise while predicting nonlinear processes in the form of chaotic behavior. The evaluated models include the perceptron neural network using backpropagation (BPN), the recurrent neural network (RNN) and case based reasoning (CBR). The concepts are crystallized through a case study in predicting a Henon process in the presence of various patterns of noise.

• 대한전기학회:학술대회논문집
• /
• 대한전기학회 1991년도 하계학술대회 논문집
• /
• pp.402-405
• /
• 1991

This paper represents a new method for computing reliability indices by using Large Deviation method which is one of the probabilistic production cost simulations. The reliability measures are based on the models used for the loads and for the generating unit failure states. In computing these measures it has been tacitly assumed that the values of all parameters in the models are precisely known. In fact, however, some of these values must often be chosen with a considerable degree of uncertainty involved. This is particularly true for the forecast peak loads in the load model, where there is an inherent uncertainty in the method of forecasting, which are frequently based on insufficient statistics. In this paper, the effect of load forecasting uncertainty on the LOLP(Loss of Load Probability), is investigated. By applying the Large Deviation method to the IEEE Rilability Test System, it is verified that the proposed method is generally very accurate and very fast for computing system reliability indices.

• 한국수자원학회:학술대회논문집
• /
• 한국수자원학회 2015년도 학술발표회
• /
• pp.225-225
• /
• 2015

Drought events usually evolve slowly in time and their impacts generally span a long period of time. This indicates that the sequence of drought is not completely random. The Hidden Markov Model (HMM) is a probabilistic model used to represent dependences between invisible hidden states which finally result in observations. Drought characteristics are dependent on the underlying generating mechanism, which can be well modelled by the HMM. This study employed a HMM with Gaussian emissions to fit the Standardized Precipitation Index (SPI) series and make multi-step prediction to check the drought characteristics in the future. To estimate the parameters of the HMM, we employed a Bayesian model computed via Markov Chain Monte Carlo (MCMC). Since the true number of hidden states is unknown, we fit the model with varying number of hidden states and used reversible jump to allow for transdimensional moves between models with different numbers of states. We applied the HMM to several stations SPI data in South Korea. The monthly SPI data from January 1973 to December 2012 was divided into two parts, the first 30-year SPI data (January 1973 to December 2002) was used for model calibration and the last 10-year SPI data (January 2003 to December 2012) for model validation. All the SPI data was preprocessed through the wavelet denoising and applied as the visible output in the HMM. Different lead time (T= 1, 3, 6, 12 months) forecasting performances were compared with conventional forecasting techniques (e.g., ANN and ARMA). Based on statistical evaluation performance, the HMM exhibited significant preferable results compared to conventional models with much larger forecasting skill score (about 0.3-0.6) and lower Root Mean Square Error (RMSE) values (about 0.5-0.9).

• 전기전자학회논문지
• /
• 제23권1호
• /
• pp.127-133
• /
• 2019

미래에 스마트 그리드 도입을 위해 전력수요예측은 중요한 연구 분야 중 하나이다. 하지만 전력데이터는 많은 외부적 요소들에 영향을 받기 때문에 예측하기 어렵다. 기존의 전력수요예측 방법들은 가공되지 않은 전력데이터를 그대로 이용하기 때문에 정확도 높은 예측을 하는데 한계가 있어왔다. 본 논문에서는 가공되지 않은 전력데이터를 이용하는 전력수요예측의 문제를 해결하기 위해 확률기반 학습알고리즘을 제안한다. 확률 모델은 전력데이터의 확률적 특성을 분석하기에 적합하다. 제안한 모델의 중기 전력수요예측 성능을 비교하기 위해 신경망 네트워크 중 하나인 순환신경망과 성능 비교를 해보았다. 매사추세츠 대학에서 제공한 전력데이터를 이용하여 성능 비교를 한 결과 본 논문에서 제안한 확률기반 학습알고리즘이 중기 수요예측에 더 좋은 성능을 나타냄을 확인하였다.

• 한국수자원학회:학술대회논문집
• /
• 한국수자원학회 2015년도 학술발표회
• /
• pp.243-243
• /
• 2015

Low-flow simulation and forecasting is one of the emerging issues in hydrology due to the increasing demand of water in dry periods. Even though low-flow simulation and forecasting remains a difficult issue for hydrologists better simulation and earlier prediction of low flows are crucial for efficient water management. The UN has never stated that South Korea is in a water shortage. However, a recent study by MOLIT indicates that Korea will probably lack water by 4.3 billion m3 in 2020 due to several factors, including land cover and climate change impacts. The two main situations that generate low-flow events are an extended dry period (summer low-flow) and an extended period of low temperature (winter low-flow). This situation demands the hydrologists to concentrate more on low-flow hydrology. Korea's annual average precipitation is about 127.6 billion m3 where runoff into rivers and losses accounts 57% and 43% respectively and from 57% runoff discharge to the ocean is accounts 31% and total water use is about 26%. So, saving 6% of the runoff will solve the water shortage problem mentioned above. The main objective of this study is to present the hydrological modelling approach for low-flow simulation and forecasting using a model that have a capacity to represent the real hydrological behavior of the catchment and to address the water management of summer as well as winter low-flow. Two lumped hydrological models (GR4J and CAT) will be applied to calibrate and simulate the streamflow. The models will be applied to Seolmacheon catchment using daily streamflow data at Jeonjeokbigyo station, and the Nash-Sutcliffe efficiencies will be calculated to check the model performance. The expected result will be summarized in a different ways so as to provide decision makers with the probabilistic forecasts and the associated risks of low flows. Finally, the results will be presented and the capacity of the models to provide useful information for efficient water management practice will be discussed.

• 해양환경안전학회지
• /
• 제24권3호
• /
• pp.302-310
• /
• 2018

Despite the accident rate for fishing vessels accounts for 70% of all maritime accidents, few studies on such accidents have been done and most of the them mainly focus on causes and mitigation policies to reduce that accident rate. Thus, this risk analysis on sea accidents is the first to be performed for the successful and efficient implementation of accident reducing measures. In risk analysis, risk is calculated based on the combination of frequency and the consequence of an accident, and is usually expressed as a single number. However, there exists uncertainty in the risk calculation process if one uses a limited number of data for analysis. Therefore, in the study we propose a probabilistic simulation method to forecast risk not as a single number, but in a range of possible risk values. For the capability of the proposed method, using the criteria with the ALARP region, we show the possible risk values spanning across the different risk regions, whereas the single risk value calculated from the existing method lies in one of the risk regions. Therefore, a decision maker could employ appropriate risk mitigation options to handle the risks lying in different regions. For this study, we used fishing vessel accident data from 1988 to 2016.

• Water Engineering Research
• /
• 제5권2호
• /
• pp.101-109
• /
• 2004

Aggregating information by combining forecasts from two or more forecasting methods is an alternative to using forecasts from just a single method to improve forecast accuracy. This paper describes the development and use of a monthly inflow forecast model based on an optimal linear combination (OLC) of forecasts derived from naive, persistence, and Ensemble Streamflow Prediction (ESP) forecasts. Using the cross-validation technique, the OLC model made 1-month ahead probabilistic forecasts for the Chungju multi-purpose dam inflows for 15 years. For most of the verification months, the skill associated with the OLC forecast was superior to those drawn from the individual forecast techniques. Therefore this study demonstrates that OLC can improve the accuracy of the ESP forecast, especially during the dry season. This study also examined the value of the OLC forecasts in reservoir operations. Stochastic Dynamic Programming (SDP) derived the optimal operating policy for the Chungju multi-purpose dam operation and the derived policy was simulated using the 15-year observed inflows. The simulation results showed the SDP model that updated its probability from the new OLC forecast provided more efficient operation decisions than the conventional SDP model.