• Journal of Energy Engineering
• /
• v.6 no.1
• /
• pp.68-76
• /
• 1997

This paper presents a new neural network training algorithm which reduces the required training time considerably and overcomes many of the shortcomings presented by the conventional back-propagation algorithm. The algorithm uses a modified form of the back-propagation algorithm to minimize the mean squared error between the desired and actual outputs with respect to the inputs to the nonlinearities. Artificial Neural Network (ANN) model using the new algorithm is applied to forecast the short-term electric load. Inputs to the ANN are past loads and the output of the ANN is the hourly load forecast for a given day.

• Journal of Industrial Technology
• /
• v.31 no.A
• /
• pp.3-9
• /
• 2011

Simulation plays an important role for system analysis. In this study, a manufacturing system has been analyzed through computer simulation. We first briefly explain the considered system with prevailing problems. We then build a simulation model using ARENA simulation language. Based on two selected performance measures, material transporter load and hourly throughput, explicit system analyses have been performed. We addressed four parameters - variation of the processing time, number of raw material transporters, quality failure rate, and machine failures - as the input parameters affecting the output measures selected. We adopted Taguchi's orthogonal array in statistical experimental design and drew meaningful results from the analysis. The results given in the study may provide a good guidance for practical applications.

• Korean Journal of Remote Sensing
• /
• v.19 no.5
• /
• pp.341-349
• /
• 2003

The so-called B-method, a simplified surface energy budget, permits calculation of daily actual evapotranspiration (ET) using remotely sensed data, such as NOAA-AVHRR. Even if the use of satellite data allows estimation of the albedo and surface temperature, this model requires meteorological data measured at ground-level to obtain the other inputs. In addition, a difficulty may be occurred by the difference of temporal scales between the net radiation in daily scale and instantaneous measurement at midday of the surface and air temperatures because the data covered whole day are necessary to obtain accumulated daily net radiation. In order to solve these problems, this study attempted a modification of B-method through an extension of hourly ET value calculated using a complete instantaneous inputs. The estimation of the daily apparent ET from newly proposed system showed a root mean square error of 0.26 mm/day as compared the output obtained from the classical model. It is evident that this may offer more rapid estimation and reduced data volume.

• Atmosphere
• /
• v.24 no.3
• /
• pp.343-364
• /
• 2014

This study presents local and synoptic conditions associated with extreme heavy snowfall events in the Yeongdong region, as well as the temporal and spatial variability of these conditions. During the last 12 years (2001~2012), 3 extreme snowfall events occurred in the Yeongdong region, which recorded daily snowfall greater than 50 cm, respectively. In these events, one of the noticeable features is the occurrence of heavy hourly snowfall greater than 10 cm. It was reported from satellite analysis that these heavy snowfall may be closely related to mesoscale convective clouds. In this paper the 3 extreme events are examined on their synoptic environments associated with the developments of mesoscale convective system using numerical model output. These 3 events all occurred in strongly forced synoptic environments where 500 and 300 hPa troughs and 500 hPa thermal troughs were evident. From the analysis of diagnostic variables, it was found in all 3 events that absolute vorticity and cold air advection were dominant in the Yeongdong region and its surrounding sea at upper levels, especially at around 500 hPa (absolute vorticity: $20{\sim}60{\times}10^{-5}s^{-1}$, cold air advection: $-10{\sim}-20^{\circ}C$ $12hr^{-1}$). Moreover, the spatial distributions of cold advection showed mostly the shape of a narrow band along the eastern coast of Korea. These features of absolute vorticity and cold advection at 500 hPa were sustained for about 10 hours before the occurrence of maximum hourly snowfall.

• Proceedings of the Korea Water Resources Association Conference
• /
• 2023.05a
• /
• pp.246-246
• /
• 2023

Multi-purpose dams are operated accounting for both physical and socioeconomic factors. This study aims to evaluate the utility of a deep learning algorithm-based model for three multi-purpose dam operation (Seomjin River dam, Juam dam, and Juam Control dam) in Seomjin River. In this study, the Gated Recurrent Unit (GRU) algorithm is applied to predict hourly water level of the dam reservoirs over 2002-2021. The hyper-parameters are optimized by the Bayesian optimization algorithm to enhance the prediction skill of the GRU model. The GRU models are set by the following cases: single dam input - single dam output (S-S), multi-dam input - single dam output (M-S), and multi-dam input - multi-dam output (M-M). Results show that the S-S cases with the local dam information have the highest accuracy above 0.8 of NSE. Results from the M-S and M-M model cases confirm that upstream dam information can bring important information for downstream dam operation prediction. The S-S models are simulated with altered outflows (-40% to +40%) to generate the simulated water level of the dam reservoir as alternative dam operational scenarios. The alternative S-S model simulations show physically inconsistent results, indicating that our deep learning algorithm-based model is not explainable for multi-purpose dam operation patterns. To better understand this limitation, we further analyze the relationship between observed water level and outflow of each dam. Results show that complexity in outflow-water level relationship causes the limited predictability of the GRU algorithm-based model. This study highlights the importance of socioeconomic factors from hidden multi-purpose dam operation processes on not only physical processes-based modeling but also aritificial intelligence modeling.

• Korean Journal of Soil Science and Fertilizer
• /
• v.44 no.6
• /
• pp.991-997
• /
• 2011

Soil erosion and sediment has been known as one of pollutants causing water quality degradation in water bodies. With global warming issues worldwide, various soil erosion studies have been performed. Although on-site monitoring of sediment loss would be an ideal method to evaluate soil erosion condition, modeling approaches have been utilized to estimate soil erosion and to evaluate various best management practices on soil erosion reduction. Although the USLE has been used in soil erosion estimation for the last 40 years, the USLE model has limitations in estimating event-based soil erosion reflecting rainfall intensity and rainfall duration for long-term period. Thus, the calibrated model, capable of simulating soil erosion using hourly rainfall data, was utilized in this study to evaluate the effects of rainfall amount and rainfall intensity on soil erosion. It was found that USLE soil erosion value is $3.06ton\;ha^{-1}\;yr^{-1}$, while soil erosion values from 2006~2010 were $2.469ton\;ha^{-1}\;yr^{-1}$, $0.882ton\;ha^{-1}\;yr^{-1}$, $1.489ton\;ha^{-1}\;yr^{-1}$, $2.158ton\;ha^{-1}\;yr^{-1}$, $1.602ton\;ha^{-1}\;yr^{-1}$, respectively. Especially, soil erosion from single storm event for 2008-2010 would be responsible for 30% or more of annual soil loss. As shown in this study, hourly soil erosion estimation system would provide more detailed output from the study area. In addition, the effects of rainfall intensity on soil erosion could be evaluated with this system.

• 한국태양에너지학회:학술대회논문집
• /
• 2008.04a
• /
• pp.49-54
• /
• 2008

This study is to investigate an optimal size and position of PV system for apartment complex enough the electrical load matching analysis. The 4 types of arrangements of apartment buildings are considered as follows; ㅡtype, alternative ㅡtype, ㄱtype and ㅁtype. We assume that the studied site is composed of 9 buildings. Firstly, solar access evaluation of roof and facade in apartment buildings was performing with the hourly simulations of total received insolation on each surface considering the hading effect of buildings. Electrical load profile of typical Korean apartments was investigated for the lad matching analysis. To calculate an annual total PV output, we used MERIT program which is a sourly based load matching tool developed by ESRU. TRY weather data of Daejeon are applied for this analysis. Result shows that approximately 11% of total electric load of the site can be supplied by the PV system in the case of full installation of PV system at the whole south-face roof area of 9 buildings in this stuided apartment complex. Depending of a various installation option of roof and facade area, the possible ratio of PV supply in total electrical load varies from 9% to 42%. Among the 4 arrangement types, the ㅡ type revealed the best option for the maximum output of PV system.

• Journal of the Korean Solar Energy Society
• /
• v.39 no.6
• /
• pp.113-125
• /
• 2019

With the opening of the small power brokerage business market in December 2018, the small power trading market has started in Korea. Operators must submit the day-ahead estimates of power output and receive incentives based on its accuracy. Therefore, the accuracy of power generation forecasts is directly affects profits of the operators. The forecasting process for power generation can be divided into two procedure. The first is to forecast solar irradiation and the second is to transform forecasted solar irradiation into power generation. There are two methods for transformation. One is to simulate with physical model, and another is to use regression model. In this study, we found the best-fit regression model by analyzing hourly data of PV output and solar irradiation data during three years for 242 PV plants in Korea. The best model was not a linear model, but a sigmoidal model and specifically a Gompertz model. The combined linear regression and Gompertz curve was proposed because a the curve has non-zero y-intercept. As the result, R² and RMSE between observed data and the curve was significantly reduced.

• Journal of Sensor Science and Technology
• /
• v.17 no.1
• /
• pp.81-85
• /
• 2008

This paper addressed real-time urine monitoring device for intensive care patients. The device was developed to detect and count each urine drop using optical sensor and calculate the current urine output volume and its hourly rate. In experiment, the water volume scale of drainage bottle was observed and compared with the count of the device so that the volume of each drop was found to vary with the dropping rate per minute. From this measurement, the relationship equation was derived to estimate the total water volume from the drop rate (correlation coefficient : r= 0.99). The developed device could be applied to count patient's urine drop successfully. Therefore, this device can be used to monitor intensive care patient's urine status in real-time.

• Asian-Australasian Journal of Animal Sciences
• /
• v.4 no.1
• /
• pp.79-83
• /
• 1991

Diurnal variation in faecal chromic oxide levels was estimated from 4 hourly faecal sampling of 56 wethers allotted to one of six different feeding patterns. Sampling occurred on days 6 and 7 (Period 1) after a controlled release device was administered, and was repeated on days 14 and 15 (Period 2) following re-randomisation of the wethers to the feeding patterns. Increasing the frequency of feeding tended to be associated with higher faecal chromic oxide levels (p < 0.05), particularly in wethers fed thrice daily at 8 hour intervals. There was no interaction between feeding pattern and period, sampling day or time within-day. There were significant period x time within-day (p < 0.05) and day x time within-day (p < 0.001) interactions, indicating that variation in faecal chromic oxide between sample times was not consistent. This implies that sampling at any time of day is unlikely to result in a biased estimate of pasture intake, providing sufficient samples are collected. Significant period (p < 0.001) and period x day (p < 0.01) effects were associated with slow faecal chromic oxide equilibration in period 1. Equilibration did not occur until after day 7, indicating a need for caution when commencing sampling.