http://chineseinput.net/에서 pinyin(병음)방식으로 중국어를 변환할 수 있습니다.
변환된 중국어를 복사하여 사용하시면 됩니다.
권오익(Oh Ik Kwon),김영일(Young Il Kim),김선혜(Sean Hay Kim) 대한설비공학회 2021 설비공학 논문집 Vol.33 No.3
Quantification of building energy performance is required for diagnosis, prediction, and evaluation of energy efficient use and performance improvement of buildings. Implementing a resistance capacity (RC) model, this study presents a hydrodynamic physical model that can predict the distribution of room temperatures with sensors of minimum quantity. The proposed prediction model could predict indoor thermal behavior very similarly, even though it did not link with other models or apply optimization techniques during the verification process. Under conditions not affected by thermodynamic parameters, the root mean square error (RMSE) range for predicting air temperature of other spaces was 0.115℃ to 0.357℃ with an average indoor air temperature of 0.132℃. Predictive models with simple input conditions developed in this study could be integrated with other various models and used for optimal control of building energy.
그린리모델링 사전진단을 위한 실내온도 센서기반 건물에너지 예측 모델
권오익(Oh Ik Kwon),이인혜(In Hye Lee),연창근(Chang kun Yeon),김영일(Young Il Kim) 대한설비공학회 2022 대한설비공학회 학술발표대회논문집 Vol.2022 No.6
The building energy model for energy efficiency of existing buildings must be simple and accurate. In general, white box models take a lot of time and money to obtain parameter information, but nevertheless, the analysis results differ from the actual energy consumption. The black box model is widely used because it is estimated based on actual performance data of the building and provides excellent accuracy. Linear regression is a popular method among them. The method of linear regression of outdoor temperature and energy use has been frequently used for a long time. However, this approach requires a lot of accumulated data. Also, it is not possible to consider the indoor environment of the building. To improve this problem, propose a delta-T-based building energy model using a indoor temperature sensor. The proposed model had better energy usage prediction performance than other models using outdoor temperatures, even considering systematic errors in sensors. It plans to further verify the model after the green remodeling of the target building is completed. The proposed model is available in the following situations: (1) A Comparative Analysis of Energy Efficiency Levels in the Operating Stage of Existing Buildings (2) Quantitative Analysis of Energy Performance Improvement Effect of Existing Buildings (3) Support for real-time indoor environment and building energy management monitoring in connection with low-cost sensor network for buildings without building energy management system
권오익(Oh Ik Kwon),엄태윤(Tae Yun Aum),김영일(Young Il Kim) 대한설비공학회 2022 대한설비공학회 학술발표대회논문집 Vol.2022 No.11
In this study, a machine learning model for predicting building energy consumption was compared and analyzed. The models to be compared are Multiple Linear Regression (MLR), Support Vector Machine (SVM), Gaussian Process Regression (GPR), Artificial Neural Networks (ANN), and Decision Tree (DT). As a result of the analysis, ANN showed the best prediction performance with RMSE 0.212. GPR was RMSE 0.213, showing similar but slightly lower predictive performance. MLR and DT, which are easy to model, showed RMSE 0.320 and 0.284, respectively. ANN, GPR, and SVM, which are relatively difficult to model, showed better predictive performance than MLR and DT, which are relatively easy to model. It is confirmed that the predictive performance of the machine learning model is influenced by the dataset and does not provide an understanding of the predictive process, but can work effectively for predicting building energy consumption. Furthermore, in order to expand the usability of the engineering aspect of the machine learning model, it is necessary to have a building energy model with a structure that enables the convenience of model creation and use, and selection and handling of physical parameters.
그린리모델링 사전진단을 위한 실내온도 센서기반 건물에너지 예측 모델
권오익(Oh Ik Kwon),이인혜(In Hye Lee),연창근(Chang kun Yeon),김영일(Young Il Kim) 대한설비공학회 2022 대한설비공학회 학술발표대회논문집 Vol.2022 No.6
The building energy model for energy efficiency of existing buildings must be simple and accurate. In general, white box models take a lot of time and money to obtain parameter information, but nevertheless, the analysis results differ from the actual energy consumption. The black box model is widely used because it is estimated based on actual performance data of the building and provides excellent accuracy. Linear regression is a popular method among them. The method of linear regression of outdoor temperature and energy use has been frequently used for a long time. However, this approach requires a lot of accumulated data. Also, it is not possible to consider the indoor environment of the building. To improve this problem, propose a delta-T-based building energy model using a indoor temperature sensor. The proposed model had better energy usage prediction performance than other models using outdoor temperatures, even considering systematic errors in sensors. It plans to further verify the model after the green remodeling of the target building is completed. The proposed model is available in the following situations: (1) A Comparative Analysis of Energy Efficiency Levels in the Operating Stage of Existing Buildings (2) Quantitative Analysis of Energy Performance Improvement Effect of Existing Buildings (3) Support for real-time indoor environment and building energy management monitoring in connection with low-cost sensor network for buildings without building energy management system
권오익(Oh Ik Kwon),엄태윤(Tae Yun Aum) 대한설비공학회 2020 대한설비공학회 학술발표대회논문집 Vol.2020 No.6
This study reviewed the validity of dynamic simulation for domestic A/C capacity selection with the aim of developing evaluation tools for A/C capacity selection and rational energy management. Through dynamic simulation, the cooling energy consumption and operation status of the facility were identified, and the method of selecting facility capacity of KS C 9306 and ASHRAE 90.1 were compared and analyzed. The A/C capacity was selected according to KS C 9306 and simulated using Energyplus and DesignBuilder. Dynamic simulation results show that if A/C capacity is insufficient due to lack of calibration thermal load considerations, residential and non-residential analyses show that cooling energy consumption is reduced by 5.5 to 12.1%, respectively, and cooling load is not handled due to lack of system capacity. In addition, the results of determining the adequacy of A/C capacity were analyzed equally in the KS C 9306 and ASHRAE 90.1 UMLH criteria. The usefulness of dynamic simulation has been identified, and the generalization of the study results requires a review of the type and size of the building required for dynamic simulation, system diversity and profile definition.
권오익(Oh Ik Kwon),김영일(Young Il Kim),김선혜(Sean Hay Kim) 대한설비공학회 2018 대한설비공학회 학술발표대회논문집 Vol.2018 No.11
The purpose of this study is to compare the change of external surface convection heat transfer coefficient based on wind speed change by building height. At a height of 10 m, wind speed is 3.5 m/s and external surface heat Transfer Coefficient is 22.59 W/(㎡·K). At 100 m height, wind speed is 7.5 m/s and external surface heat Transfer coefficient is 38.52 W/(㎡·K). At a height of 500 m, the wind speed is 12.7 m/s and the external surface heat Transfer coefficient is 59.50 W/(㎡·K). This difference in external surface heat transfer Coefficient does not change the building energy efficiency rating. Considering the long life cycle of the building, it is necessary to improve the insulation performance according to the height of the building.