• 설비공학논문집
• /
• 제20권3호
• /
• pp.205-212
• /
• 2008

This paper describes the application of an inverse building model to a calibrated forward building model using EnergyPlus program. Typically, inverse models are trained using measured data. However, in this study, an inverse building model was trained using data generated by an EnergyPlus model for an actual office building. The EnergyPlus model was calibrated using field data for the building. A training data set for a month of July was generated from the EnergyPlus model to train the inverse model. Cooling load prediction of the trained inverse model was tested using another data set from the EnergyPlus model for a month of August. Predicted cooling loads showed good agreement with cooling loads from the EnergyPlus model with root-mean square errors of 4.11%. In addition, different control strategies with dynamic cooling setpoint variation were simulated using the inverse model. Peak cooling loads and daily cooling loads were compared for the dynamic simulation.

• International Journal of Computer Science & Network Security
• /
• 제22권7호
• /
• pp.301-307
• /
• 2022

Energy consumption has grown alongside dramatic population increases. Statistics show that buildings in particular utilize a significant amount of energy, worldwide. Because of this, building energy prediction is crucial to best optimize utilities' energy plans and also create a predictive model for consumers. To improve energy prediction performance, this paper proposes a ResNet-LSTM model that combines residual networks (ResNets) and long short-term memory (LSTM) for energy consumption prediction. ResNets are utilized to extract complex and rich features, while LSTM has the ability to learn temporal correlation; the dense layer is used as a regression to forecast energy consumption. To make our model more robust, we employed Huber loss during the optimization process. Huber loss obtains high efficiency by handling minor errors quadratically. It also takes the absolute error for large errors to increase robustness. This makes our model less sensitive to outlier data. Our proposed system was trained on historical data to forecast energy consumption for different time series. To evaluate our proposed model, we compared our model's performance with several popular machine learning and deep learning methods such as linear regression, neural networks, decision tree, and convolutional neural networks, etc. The results show that our proposed model predicted energy consumption most accurately.

• 한국원자력학회:학술대회논문집
• /
• 한국원자력학회 1997년도 춘계학술발표회논문집(2)
• /
• pp.407-412
• /
• 1997

The predictive results between a dynamic food-chain model (DYNACON) and an equilibrium model (NRC model) were compared to show the physical validity of DYNACON. Although the mathematical formulations and transport processes of radionuclides in the environment are different between two models, the comparative study shows good agreement for deposition events that occur during the growing season of plants.

• 전기학회논문지
• /
• 제60권6호
• /
• pp.1103-1111
• /
• 2011

This paper is developed to Energy Balance Flow show the flow of total energy resource be used nationally. The Energy Balance Flow is applicable of demand management factor through the analysis of foreign energy model of supply and demand and energy statistic data in the country. This study is based on and developed to Energy system management model is able to appraisal efficient of energy cost cutting, CO2 emission reduction and Energy saving at the national level calculated effect reached amount of primary energy to change of energy flow followed application of demand side management factor is able to appraisal quantitatively at the total energy to model of demand and supply.

• 국제학술발표논문집
• /
• The 9th International Conference on Construction Engineering and Project Management
• /
• pp.1256-1263
• /
• 2022

With the advent of the Fourth Industrial Revolution, the amount of energy used in buildings has been increasing due to changes in the energy use structure caused by the massive spread of information-oriented equipment, climate change and greenhouse gas emissions. For the efficient use of energy, it is necessary to have a plan that can predict and reduce the amount of energy use according to the type of energy source and the use of buildings. To address such issues, this study presents a model embedded in a digital twin that predicts energy use in buildings. The digital twin is a system that can support a solution of urban problems through the process of simulations and analyses based on the data collected via sensors in real-time. To develop the energy use prediction model, energy-related data such as actual room use, power use and gas use were collected. Factors that significantly affect energy use were identified through a correlation analysis and multiple regression analysis based on the collected data. The proof-of-concept prototype was developed with an exhibition facility for performance evaluation and validation. The test results confirm that the error rate of the energy consumption prediction model decreases, and the prediction performance improves as the data is accumulated by comparing the error rates of the model. The energy use prediction model thus predicts future energy use and supports formulating a systematic energy management plan in consideration of characteristics of building spaces such as the purpose and the occupancy time of each room. It is suggested to collect and analyze data from other facilities in the future to develop a general-purpose energy use prediction model.

• 한국연소학회:학술대회논문집
• /
• 한국연소학회 2004년도 제28회 KOSCO SYMPOSIUM 논문집
• /
• pp.158-164
• /
• 2004

In the present work, the numerical model was refined to predict the thermal analysis of energy storage in a fixed beds during charging mode. The governing energy equations of both fluid and the solid particles along with their initial and boundary conditions are derived using a two-phase, one dimensional model. The refined model is carried out by taking into account change of (air density , air specific heat) with air temperature and also by taking into considerations heat losses from bed to surrounding. Finite difference method was used to obtain solution of two governing energy equations of both fluid and solid particles through a computer program especially constructed for this purpose. The temperature field for the air and the solid are obtained, also energy stored inside the bed is computed. A comparison between refined model and non refined model is done. Finally using refined model the effect of bed material (Glass, Fine clay ,and aluminum ), and air flow rate per unit area Ga (0.3, 0.4, and 0.5 kg/$m^2$-s) on energy storage characteristics was studied.

• International Journal of Air-Conditioning and Refrigeration
• /
• 제12권4호
• /
• pp.169-175
• /
• 2004

Case study was performed to analyze energy load for department stores and develop energy load model to be applied to a cogeneration system. Energy loads of 14 depart­ments were analyzed based on energy load sheets written by operators and energy load of one department store was measured through modem communication for a year. Energy load of department stores showed various trends depending on when they were opened or closed, or by hour, day and month. In this paper, the measurement was compared with the data in energy load sheets and resolved, and energy load model for a department store was built. It is important to use an accurate energy load model for an accurate feasibility study applying a cogeneration system to buildings.

• Journal of Electrical Engineering and Technology
• /
• 제11권2호
• /
• pp.470-479
• /
• 2016

The model based method for energy saving of the separately excited DC motor drive system is proposed in the paper. The accurate power loss model is necessary for this method. Therefore, the adaptive tabu search algorithm is applied to identify the parameters in the power loss model. The field current values for minimum power losses at any load torques and speeds are calculated by the proposed method. The rule based controller is used to control the field current and speed of the motor. The experimental results confirm that the model based method can successfully provide the energy saving for separately excited DC motor drive. The maximum value of the energy saving is 48.61% compared with the conventional drive method.

• 한국초전도ㆍ저온공학회논문지
• /
• 제21권3호
• /
• pp.6-12
• /
• 2019

Extended Drude model has been used to obtain information of correlations from measured optical spectra of strongly correlated electron systems. The optical self-energy can be defined by the extended Drude model formalism. One can extract the optical self-energy and the electron-boson spectral density function from measured reflectance spectra using a well-developed usual process, which is consistent with several steps including the extended Drude model and generalized Allen's formulas. Here we used a reverse process of the usual process to investigate the extended Drude analysis when an additional low-energy interband transition is included. We considered two typical electron-boson spectral density model functions for two different (normal and d-wave superconducting) material states. Our results show that the low-energy interband transition might give significant effects on the electron-boson spectral density function obtained using the usual process. However, we expect that the low-energy interband transition can be removed from measured spectra in a proper way if the transition is well-defined or well-known.

• 한국표면공학회지
• /
• 제43권3호
• /
• pp.159-164
• /
• 2010

Surface roughness of deposited or etched film strongly depends on ion bombardment. Relationships between ion bombardment variables and surface roughness are too complicated to model analytically. To overcome this, an empirical neural network model was constructed and applied to a deposition process of silicon nitride (SiN) films. The films were deposited by using a pulsed plasma enhanced chemical vapor deposition system in $SiH_4$-$NH_4$ plasma. Radio frequency source power and duty ratio were varied in the range of 200-800 W and 40-100%. A total of 20 experiments were conducted. A non-invasive ion energy analyzer was used to collect ion energy distribution. The diagnostic variables examined include high (or) low ion energy and high (or low) ion energy flux. Mean surface roughness was measured by using atomic force microscopy. A neural network model relating the diagnostic variables to the surface roughness was constructed and its prediction performance was optimized by using a genetic algorithm. The optimized model yielded an improved performance of about 58% over statistical regression model. The model revealed very interesting features useful for optimization of surface roughness. This includes a reduction in surface roughness either by an increase in ion energy flux at lower ion energy or by an increase in higher ion energy at lower ion energy flux.