• Journal of Institute of Convergence Technology
• /
• v.9 no.1
• /
• pp.31-36
• /
• 2019

After 30 years of KKP model analysis and extended 30 years of accuracy analysis, the unique correlation and various problems between measured horizontal surface insolation and measured weather data are found in this paper. The KKP model's 10yrs daily total horizontal surface insolation forecasting was averaged about 97.7% on average, and the forecasting accuracy at peak times per day was about 92.1%, which is highly applicable regardless of location and weather conditions nationwide. The daily total solar radiation forecasting accuracy of the modified KKP cloud model was 98.9%, similar to the KKP model, and 93.0% of the forecasting accuracy at the peak time per day. And the results of evaluating the accuracy of calculation for 30 years of KKP model were cloud model 107.6% and cloud model 95.1%. During the accuracy analysis evaluation, this study found that inaccuracies in measurement data of cloud cover should be clearly assessed by the Meteorological Administration.

• Proceedings of the KIEE Conference
• /
• 2000.11a
• /
• pp.184-186
• /
• 2000

This paper proposes the introduction of TAR(Threshold Autoregressive) model for short-term load forecasting including temperature variable. TAR model is a piecewise linear autoregressive model. In the scatter diagram of daily peak load versus daily maximum or minimum temperature, we can find out that the load-temperature relationship has a negative slope in lower regime and a positive slope in upper regime due to the heating and cooling load, respectively. In this paper, daily peak load was forecasted by applying TAR model using this load-temperature characteristic in these regimes. The results are compared with those of linear and quadratic regression models.

• Proceedings of the KIEE Conference
• /
• 1995.07b
• /
• pp.578-580
• /
• 1995

In this paper, a method for the daily maximum load forecasting which uses a chaotic time series in power system and artificial neural network. We find the characteristics of chaos in power load curve and then determine a optimal embedding dimension and delay time, For the load forecast of one day ahead daily maximum power, we use the time series load data obtained in previous year. By using of embedding dimension and delay time, we construct a strange attractor in pseudo phase plane and the artificial neural network model trained with the attractor font mentioned above. The one day ahead forecast errors are about 1.4% of absolute percentage average error.

• Magazine of the Korean Society of Agricultural Engineers
• /
• v.36 no.1
• /
• pp.63-72
• /
• 1994

The objective of the study is to develop weather generators for daily rainfall and small pan evaporation and to test the applicability with recorded data. Daily rainfall forecasting model(DRFM) was developed that uses a first order Markov chain to describe rainfall seque- nces and applies an incomplete Gamma function to predict the amount of precipitation. Daily evaporation forecasting model(DEFM) that adopts a normal distribution function to generate the evaporation for dry and wet days was also formulated. DRFM and DEFM were tested with twenty year weather data from eleven stations using Chi-square and Kolmogorov and Smirnov goodness of fit tests. The test results showed that the generated sequences of rainfall occurrence, amount of rainfall, and pan evaporation were statistically fit to recorded data from eleven, seven, and seven stations at the 5% level of significance. Generated rainfall data from DRFM were very close in frequency distri- bution patterns to records for stations all over the country. Pan evaporation for rainy days generated were less accurate than that for dry days. And the proposed models may be used as tools to provide many mathematical models with long-term daily rainfall and small pan evaporation data. An example is an irrigation scheduling model, which will be further detailed in the paper.

• The Transactions of the Korean Institute of Electrical Engineers A
• /
• v.53 no.10
• /
• pp.529-535
• /
• 2004

Load forecasting is essential in the electricity market for the participants to manage the market efficiently and stably. A wide variety of techniques/algorithms for load forecasting has been reported in many literatures. These techniques are as follows: multiple linear regression, stochastic time series, general exponential smoothing, state space and Kalman filter, knowledge-based expert system approach (fuzzy method and artificial neural network). These techniques have improved the accuracy of the load forecasting. In recent 10 years, many researchers have focused on artificial neural network and fuzzy method for the load forecasting. In this paper, we propose an algorithm of a hybrid load forecasting method using fuzzy linear regression and general exponential smoothing and considering the sensitivities of the temperature. In order to consider the lower load of weekends and Monday than weekdays, fuzzy linear regression method is proposed. The temperature sensitivity is used to improve the accuracy of the load forecasting through the relation of the daily load and temperature. And the normal load of weekdays is easily forecasted by general exponential smoothing method. Test results show that the proposed algorithm improves the accuracy of the load forecasting in 1996.

• Journal of the Korea Convergence Society
• /
• v.10 no.1
• /
• pp.7-12
• /
• 2019

The application of neural networks to stock forecasting has received a great deal of attention because no assumption about a suitable mathematical model has to be made prior to forecasting and they are capable of extracting useful information from data, which is required to describe nonlinear input-output relations of stock forecasting. The paper builds neural network models to forecast daily KOrea composite Stock Price Index (KOSPI), and their performance is demonstrated. MAPEs of NN1 model show 0.427 and 0.627 in its learning and test, respectively. Based on the predicted KOSPI price, the paper proposes an alpha trading for trades in Exchange Traded Funds (ETFs) that fluctuate with the KOSPI200. The alpha trading is tested with data from 125 trade days, and its trade return of 7.16 ~ 15.29 % suggests that the proposed alpha trading is effective.

• The Korean Journal of Applied Statistics
• /
• v.32 no.1
• /
• pp.69-81
• /
• 2019

We consider bivariate long range dependent (LRD) time series forecasting using a deep learning method. A long short-term memory (LSTM) network well-suited to time series data is applied to forecast bivariate time series; in addition, we compare the forecasting performance with bivariate fractional autoregressive integrated moving average (FARIMA) models. Out-of-sample forecasting errors are compared with various performance measures for functional MRI (fMRI) data and daily realized volatility data. The results show a subtle difference in the predicted values of the FIVARMA model and VARFIMA model. LSTM is computationally demanding due to hyper-parameter selection, but is more stable and the forecasting performance is competitively good to that of parametric long range dependent time series models.

• 한국방재학회:학술대회논문집
• /
• 2008.02a
• /
• pp.89-92
• /
• 2008

The extreme heat watch warning system(EHWWS) that Korea Meterological Administration carried out a preliminary from July 1, 2007, considered both daily maximum temperature and daily maximum heat index simultaneously. It was requested revision of the standard level of EHWWS to solve the difficulty of forecasting occurred when we were considering two parameters simultaneously and we did not considering heat index according to areas. For this, we established three type standard, such as type 1 that considered both daily maximum temperature and daily maximum heat index, Under the extreme heat day that daily minimum temperature was more than $25^{\circ}C$, type 2 that considered daily maximum temperature and type 3 that considered only daily maximum heat index and then analyzed whether these 3 types satisfies the excess mortality of the extreme heat warning or not. As a results, type 1 and 2 were more explain away excess mortality each warning step than type 3. type 2 could also apply case of not to consider heat index according to areas and had a merit for extreme heat forecasting easily because the standard was simple. Therefore we think type 2 is more suitable and reasonable standard for Korea extreme heat watch warning system(KEHWWS) than type 1. In addition, we need to develop model that exactly predicts the excess mortality will be take place during the extreme heat warning and construct KEHWWS.

• 한국데이터정보과학회:학술대회논문집
• /
• 2005.04a
• /
• pp.155-163
• /
• 2005

This paper applies Support Vector Regression (SVR) to estimate and forecast nonlinear autoregressive integrated (ARI) model of the daily exchange rates of four currencies (Swiss Francs, Indian Rupees, South Korean Won and Philippines Pesos) against U.S. dollar. The forecasting abilities of SVR are compared with linear ARI model which is estimated by OLS. Sensitivity of SVR results are also examined to kernel type and other free parameters. Empirical findings are in favor of SVR. SVR method forecasts exchange rate level better than linear ARI model and also has superior ability in forecasting the exchange rates direction in short test phase but has similar performance with OLS when forecasting the turning points in long test phase.

• Proceedings of the KIEE Conference
• /
• 2001.05a
• /
• pp.230-232
• /
• 2001

This paper presents a model for short-term load forecasting using multiple time-series. We made one-hour ahead load forecasting without classifying load data according to daily load patterns(e.g. weekday. weekend and holiday) To verify its effectiveness. the results are compared with those of neuro-fuzzy forecasting model(5). The results show that the proposed model has more accurate estimate in forecasting.