• KIPS Transactions on Software and Data Engineering
• /
• v.11 no.7
• /
• pp.307-314
• /
• 2022

Recently, researches using deep learning-based models are being actively conducted to replace statistical-based time series forecast techniques to predict electric power demand. The result of analyzing the researches shows that the performance of the LSTM-based prediction model is acceptable, but it is not sufficient for long-term regional-wide power demand prediction. In this paper, we propose a WaveNet deep learning model to predict electric power demand 24-hour-ahead with temperature data in order to achieve the prediction accuracy better than MAPE value of 2% which statistical-based time series forecast techniques can present. First of all, we illustrate a delated causal one-dimensional convolutional neural network architecture of WaveNet and the preprocessing mechanism of the input data of electric power demand and temperature. Second, we present the training process and walk forward validation with the modified WaveNet. The performance comparison results show that the prediction model with temperature data achieves MAPE value of 1.33%, which is better than MAPE Value (2.33%) of the same model without temperature data.

• Journal of the Korean Society of Industry Convergence
• /
• v.24 no.3
• /
• pp.261-269
• /
• 2021

This study was evaluated the elastic wave properties according to tension of Incoloy 825 alloy with different solution treatment temperature and aging time. Solution treatment was carried out at 700, 800, 900, and 1000 ℃ for 1 hour, and aging was carried out at 700 ℃ for 1, 5, 10, and 30 hours. As the solution treatment temperature increased, the tensile strength decreased and the elongation increased. However, as the aging time increased, the tensile strength increased and the elongation decreased. The dominant frequency decreased as the solution treatment temperature increased, but increased as the aging time increased. The dominant frequency according to the solution treatment and aging time increased as the tensile strength increased, but increased despite the decrease in elongation.

• The Sea:JOURNAL OF THE KOREAN SOCIETY OF OCEANOGRAPHY
• /
• v.10 no.1
• /
• pp.100-112
• /
• 2005

Soil temperature was measured from the surface to 40 cm depth at three stations with different heights in tidal flat of Gomso Bay, west coast of Korea, for one month in every season 2004 to examine the thermal structure and the variation. Mean temperature in surface layer was higher in summer and lower in winter than in lower layer, reflecting the seasonal variation of vertically propagating structure of temperature by heating and cooling from the tidal flat surface. Standard deviation of temperature decreased from the surface to lower layer. Periodic variations of solar radiation energy and tide mainly caused short term variation of soil temperature, which was also intermittently influenced by precipitation and wind. Time series analysis showed the power spectral energy peaks at the periods of 24, 12 and 8 hours, and the strongest peak appeared at 24 hour period. These peaks can be interpreted as temperature waves forced by variations of solar radiation, diurnal tide and interaction of both variations, respectively. EOF analysis showed that the first and the second modes resolved 96% of variation of vertical temperature structure. The first mode was interpreted as the heating antl cooling from tidal flat surface and the second mode as the effect of phase lag produced by temperature wave propagation in the soil. The phase of heat transfer by 24 hour period wave, analyzed by cross spectrum, showed that mean phase difference of the temperature wave increased almost linearly with the soil depth. The time lags by the phase difference from surface to 10, 20 and 40cm were 3.2,6.5 and 9.8 hours, respectively. Vertical thermal diffusivity of temperature wave of 24 hour period was estimated using one dimensional thermal diffusion model. Average diffusivity over the soil depths and seasons resulted in $0.70{\times}10^{-6}m^2/s$ at the middle station and $0.57{\times}10^{-6}m^2/s$ at the lowest station. The depth-averaged diffusivity was large in spring and small in summer and the seasonal mean diffusivity vertically increased from 2 cm to 10 cm and decreased from 10 cm to 40 cm. Thermal propagation speeds were estimated by $8.75{\times}10^{-4}cm/s,\;3.8{\times}10{-4}cm/s,\;and\;1.7{\times}10^{-4}cm/s$ from 2 cm to 10 cm, 20 cm and 40 cm, respectively, indicating the speed reduction with depth increasing from the surface.

• Journal of Korean Society of Coastal and Ocean Engineers
• /
• v.31 no.3
• /
• pp.115-128
• /
• 2019

In this study, a probabilistic model for the estimation of yearly workable wave condition period for offshore operations is developed. In doing so, we first hindcast the significant wave heights and peak periods off the Ulsan every hour from 2003.1.1 to 2017.12.31 based on the meteorological data by JMA (Japan Meterological Agency) and NOAA (National Oceanic and Atmospheric Administration), and SWAN. Then, we proceed to derive the long term significant wave height distribution from the simulated time series using a least square method. It was shown that the agreements are more remarkable in the distribution in line with the Modified Glukhovskiy Distribution than in the three parameters Weibull distribution which has been preferred in the literature. In an effort to develop a more comprehensive probabilistic model for the estimation of yearly workable wave condition period for offshore operations, wave height distribution over the 15 years with individual waves occurring within the unit simulation period (1 hour) being fully taken into account is also derived based on the Borgman Convolution Integral. It is shown that the coefficients of the Modified Glukhovskiy distribution are $A_p=15.92$, $H_p=4.374m$, ${\kappa}_p=1.824$, and the yearly workable wave condition period for offshore work is estimated to be 319 days when a threshold wave height for offshore work is $H_S=1.5m$. In search of a way to validate the probabilistic model derived in this study, we also carry out the wave by wave analysis of the entire time series of numerically simulated significant wave heights over the 15 years to collect every duration periods of waves the height of which are surpassing the threshold height which has been reported to be $H_S=1.5m$ in the field practice in South Korea. It turns out that the average duration period is 45.5 days from 2003 to 2017, which is very close to 46 days from the probabilistic model derived in this study.

• Journal of Navigation and Port Research
• /
• v.28 no.8
• /
• pp.745-751
• /
• 2004

Following a preceding study of Shin et al.(2004), wave fields for a month of September of 2003 are simulated based on the modified WAM cycle 4 model that enables the precise wave hindcasting with fine spatial meshes, and characteristics of extreme waves at the south coast of Korea are analyzed The accuracy of applied wave model is verified by comparing computed wave parameters and corresponding ones measured at Ieodo ocean research station. The wave hindcasting of typhoon 'Maemi' with an hour time interval reveals the extreme wave characteristics at 4 primary locations of south coast of Korea as follows: 1) At the front sea of Chaguido in the south of Jeju-do, the maximum significant wave height, mean wave period and mean wave direction appear to be 7.41m, 13.65s and $6.4^{\circ}$ respectively at 16:00 KST of Sep. 12, 2003. 2) At the entrance of Masan Bay, 12.50m, 13.65s and $1.2^{\circ}$ at 21:00 KST of Sep. 12. 3) At the front sea of Suyoung Bay, 13.85m, 13.81s and $0.2^{\circ}$ at 22;00 KST of Sep. 12. 4) At the front sea of Ulsan port, l1.00m, 13.25s and $2.8^{\circ}$ at 23:00 KST of Sep. 12.

• Ocean Systems Engineering
• /
• v.1 no.4
• /
• pp.263-283
• /
• 2011

The performance of a rig tie-down system on a TLP (Tension Leg Platform) is investigated for 10-year, 100-year, and 1000-year hurricane environments. The inertia loading on the derrick is obtained from the three-hour time histories of the platform motions and accelerations, and the dynamic wind forces as well as the time-dependent heel-induced gravitational forces are also applied. Then, the connection loads between the derrick and its substructure as well as the substructure and deck are obtained to assess the safety of the tie-down system. Both linear and nonlinear inertia loads on the derrick are included. The resultant external forces are subsequently used to calculate the loads on the tie-down clamps at every time step with the assumption of rigid derrick. The exact dynamic equations including nonlinear terms are used with all the linear and second-order wave forces considering that some dynamic contributions, such as rotational inertia, centripetal forces, and the nonlinear excitations, have not been accounted for in the conventional engineering practices. From the numerical simulations, it is seen that the contributions of the second-order sum-frequency (or springing) accelerations can be appreciable in certain hurricane conditions. Finally, the maximum reaction loads on the clamps are obtained and used to check the possibility of slip, shear, and tensile failure of the tie-down system for any given environment.

• KSCE Journal of Civil and Environmental Engineering Research
• /
• v.31 no.3D
• /
• pp.355-361
• /
• 2011

In order to properly evaluate ITS services exposed in SMART Highway project, a confident dynamic origin-destination (OD) is inevitably needed. This paper used WAVE communication information as a part of call and response (C&R) communication which constitutes core part of the technology for constructing OD. This information includes node information and vehicle information (e.g., latitude and longitude) as well as trajectory data and sample path volume date calculated using node information and vehicle information. A procedure developed to construct a dynamic OD and to validate OD is consist of 1) making toy network and one-hour 00 (random distribution), 2) collecting link information and vehicle information, 3) constructing five-minute OD, and 4) validating estimated OD result using traffic volume and travel time simultaneously. The constructed OD is about 84.79% correct within less than 20% error range for 15min traffic volume, and about 85.42%, within less than 20% error rate of 15 min travel time. Some limitations and future research agenda have also been discussed.

• Journal of Astronomy and Space Sciences
• /
• v.21 no.2
• /
• pp.101-108
• /
• 2004

The terrestrial nightglow emission in near infrared region were obtained using a Fourier Transform Spectrometer(FTS) at Esrange, Sweden ($67.90^{\circ}$N, $21.10^{\circ}$E) and the OH(4- 2) bands were used to derive temperature and airglow emission rate of the upper mesosphere. For this study, we analyzed data taken during winter of 2001/2002 and performed spectral analysis to retrieve wave information. From the Lomb-Scargle spectral analysis to the measured temperatures, dominant oscillations at various periods near tidal frequency are found. Most commonly observed waves are 4, 6, and 8 hour oscillations. Because of periods and persistence, the observed oscillations are most likely of tidal origin, i.e. zonally symmetric tides which are known to have their maximum amplitudes at the pole.

• Current Optics and Photonics
• /
• v.3 no.1
• /
• pp.33-39
• /
• 2019

We report development of a frequency-stabilized mid-infrared continuous-wave (cw) optical parametric oscillator (OPO) based on a fan-out grating MgO:PPLN crystal pumped at 1064 nm. The OPO resonator was designed as a pump-enhanced standing-wave cavity that resonates to the pump and signal beams. To realize stable operation of the OPO, we applied a modified Pound-Drever-Hall technique, which is a well-known method for powerful laser frequency stabilization. Tuning a poling period of the fan-out grating of the crystal allows wavelength-tunable OPO outputs from 1510 nm to 1852 nm and from 2500 nm to 3600 nm for signal and idler beams, respectively. At the idler wavelengths of 2500 nm, 3000 nm and 3500 nm, we achieved more than 50 mW of output powers at a pumping power of 1.1 W. The long-term stability of the OPO was confirmed by recording the power and wavelength variations of the idler for an hour.

• Journal of the Korean Society of Marine Environment & Safety
• /
• v.27 no.6
• /
• pp.701-707
• /
• 2021

In this study, we examined the sea surface temperature (SST), air temperature (AT), and their time lag in response to an extreme cold wave in 2018 and a weak cold wave in 2019, cross-correlating these to the northern wind direction frequency. The data used in this study include SST observations of seven ocean buoys Real-time Information System for Aquaculture Environment provided by the National Institute of Fisheries Science and automatic weather station AT near them recorded every hour; null data was interpolated. A finite impulse response filter was used to identify the appropriate data period. In the extreme cold wave in 2018, the seven locations indicated low SST caused by moving cold air through the northern wind direction. A warm cold wave in 2019, the locations showed that the AT data was similar to the normal AT data, but the SST data did not change notably. During the extreme cold wave of 2018, data showed a high correlation coefficient of about 0.7 and a time lag of about 14 hours between AT and SST; during the weak cold wave of 2019, the correlation coefficient was 0.44-0.67 and time lag about 20 hours between AT and SST. This research will contribute to rapid response to such climate phenomena while minimizing aquaculture damage.