To numerically investigate the flow and heat transfer characteristics of a water/Al₂O₃ nanofluid in a double-pipe helical coil heat exchanger, we simulated a two-phase Eulerian model to predict the heat transfer coefficient, Nusselt number, and pressure drop at various concentrations (i.e., volume fraction) and under diverse flow rates at the steady state. In this simulation, we used the k-epsilon turbulence model with an enhanced wall treatment method. The performance factor of the nanofluid was evaluated by accounting for the heat transfer and pressure drop characteristics. As a result, the heat transfer was enhanced by increasing the nanofluid concentration. The 1.0 vol.% nanofluid (i.e., the highest concentration) showed a heat transfer coefficient 1.43 times greater than water and a Nusselt number of 1.38 times greater than water. The pressure drop of nanofluids was greater than that of water due to the increased density and viscosity induced using nanoparticles. Based on the relationship between the Nusselt number and pressure drop, the 1.0 vol.% nanofluid was calculated to have a performance factor of 1.4 relative to water, indicating that the enhancement rate in heat transfer performance was greater than that in the pressure drop. In conclusion, the Al₂O₃ nanofluid shows potential as an enhanced working fluid in diverse heat transfer applications.

This study investigated the effect of the class number on the prediction performance of the convolutional neural network (CNN) classification model that is applied in fire detectors to reduce nuisance fire alarms by appropriately recognizing fire images including those of flames and smoke. A CNN model trained by transfer learning using five image datasets of flame, smoke, normal, haze, and light was realized and trained by altering the class number to generate the classification model. A total of three classification models were generated as follows: classification model 1 was trained using normal and fire images including flames and smoke; classification model 2 was trained using flame, smoke, and normal images; and classification model 3 was trained using flames, smoke, normal, and haze, and light images. A test image dataset independent of training was used to assess the prediction performance of the three classification models. The results indicate that the prediction accuracy for classification models 1, 2, and 3 were approximately 93.0%, 94.2%, and 97.3%, respectively. The performance of the predicted classification improved as the class number increased, because the model could learn with greater precision the features of the normal images that are similar to those of the fire images.

In this study, fire prediction performance was analyzed using convolutional neural network (CNN)-based classification models such as MobileNetV2, ResNet101, and EfficientNetB0 applicable to an edge computing-based fire detection system for improving fire safety. The fire prediction performance was evaluated using the performance evaluation measures including accuracy, recall, precision, F1-score, and the confusion matrix. The model size and inference time were assessed in terms of the light-weight classification model for the practical deployment and use. The analysis results confirmed that the EfficientNetB0 model had the highest fire prediction accuracy, and the MobileNetV2 was the best light-weight classification model. Notably, additionally learning the image features about light and haze images having similar features with those of the fire images improved the fire prediction accuracy of the light-weight MobileNetV2 model.

In this study, the energy consumption and generation characteristics, the operation status of a photovoltaic (PV) system, and the energy balance of a net-zero energy building (nZEB) in South Korea were analyzed based on the data collected over a 10-year period (2012–2021). The average annual power consumption of the nZEB was 101.3 MWh, 6.2% higher than the estimated power consumption. The PV system of the nZEB had an annual power generation capacity of 105.8 MWh, indicating an increase of 10.6% compared to the estimated value. The failure of PV systems such as module cracks, inverter failures, and performance degradation led to a decrease of 21.5% in the power generation. Energy balance analysis was conducted by comparing the energy consumption and generation data based on yearly, monthly, daily, and hourly time intervals. In addition, load coverage factor (LCF) and supply coverage factor (SCF) were used to evaluate the load matching rate. The nZEB achieved a net-zero energy status for 5 out of the 7 years of normal operation (2012–2018) based on average annual data. However, the energy balance analysis using hourly measured data showed that there was both a surplus and a shortage of power every year, and that the average annual power surplus and shortage were 56.4 MWh and 54.3 MWh, respectively. In addition, the load matching analysis showed that the annual LCF and SCF were approximately 0.36 and 0.32, respectively. Thus, the advanced nZEB design, hourly data-based energy analysis, fault diagnosis and maintenance, and the strategies enhancing the self-consumption rate should be considered to expand nZEB dissemination.