JBJunho Bang

Junho Bang

*This is an algorithmically generated author record based on publicly available information.

Current Organizations / Institutes

The organizations/institutes from the scholar's latest article


Junho Bang has a total of 18 co-authors. Here are the top co-authors by number of shared publications:


Latest Publications

  • Open Access

    Wind Resource Assessment Using Measured and ERA5 Data for Wind Turbine Performance Evaluation at a Coastal Site in South Korea

    by Somayeh Rastkar Mehrabani, Junho Bang

    This study evaluates the wind energy potential and coastal wind turbine suitability at a coastal site in South Korea using two-year measured wind data and ERA5 reanalysis data. Wind characteristics, Weibull parameters, wind power density, and turbine energy performance were analyzed using Windographer. ERA5 showed similar seasonal and directional patterns to the measured data, supporting its use for wind resource comparison and validation over the available measurement period. Among four candidate turbines, the Vestas V112-3.0 MW achieved the highest Net AEP, while the GE 1.7-100 showed the highest NCF. Finally, the CRITIC–TOPSIS method was applied using five technical and energy criteria, identifying the Vestas V112-3.0 MW as the highest-ranked under the selected technical–energy criteria with a closeness coefficient of 0.5547.

  • Open Access

    A Terrain-Structured Dual-Pareto Framework for Saemangeum Wind-Farm Re-Planning Under Measured Wind, Wake-Loss, and Cost-Chain Constraints

    by Jie Zhao, Junho Bang

    Wind-farm re-planning must reconcile wind resources, wake interactions, cable routing, foundation and installation requirements, grid access, land/water-use cost, and operation and maintenance. This study develops a terrain-structured dual-Pareto framework for the Saemangeum complex coastal site on the southwest coast of the Republic of Korea, along the inner side of its 33.9 km seawall. Stage I aggregates planning-scale points into 12 regional engineering nodes and screens them using five objectives; Stage II evaluates 43,722 layouts with 12–22 turbines inside the retained envelope. The complete 2018 wind record provides the sole full-year search and economic-evaluation basis. Under the common planning model, the selected 20-turbine, 132 MW scenario is estimated to deliver 288.7 GWh with an estimated LCOE of 184.2 USD/MWh. Relative to the 99.2 MW public proxy baseline, installed capacity rises by 33.1%, total modeled energy by 43.4%, and modeled energy per MW by about 7.8%; these are aggregate modernization-and-re-planning differences, not algorithm-only effects. In the matched eight-seed, same-technology comparison, the methods show economically comparable screening LCOE and no systematic AEP or wake-loss advantage. The dual-layer search nevertheless shortens collection cable and lowers the common terrain-adjusted CAPEX proxy in all eight seeds. The fixed layouts are also evaluated over the independently held January–July 2019 record without annualization: the main layout is estimated to deliver 159.0 GWh over 5088 valid hours, with a partial-period capacity factor of 23.7%. This held-out-period assessment does not establish annual or long-term climatic robustness.

  • Open Access

    Android-Based Real-Time Classification of Electric Fire Short-Circuit Traces Using Lightweight Deep Learning Model

    by Mohammad Hadi Nazari, Junho Bang

    This paper presents a lightweight deep learning framework for classifying electric fire short-circuit traces to enhance safety and fault diagnosis in electrical energy systems. Accurate differentiation between primary (PSCT) and secondary short-circuit traces (SSCT) is essential for identifying failure origins, yet conventional manual inspection is time-consuming and subjective. To address these limitations, we systematically evaluate three lightweight convolutional neural network (CNN) architectures MobileNetV2, MobileNetV3, and EfficientNet using transfer learning on a domain-specific image dataset. The models are assessed based on accuracy, loss, precision, recall, and F1-score. Experimental results show that EfficientNet achieves the highest classification accuracy, while MobileNetV3 demonstrates the lowest validation loss and superior generalization stability. Based on a performance–efficiency trade-off analysis, MobileNetV3 is deployed on an Android platform using TensorFlow Lite, enabling real-time, offline, and on-device inference. To the best of our knowledge, this is among the first studies to integrate lightweight CNN-based short-circuit trace classification with real-time mobile deployment for on-site energy system fault analysis. By bridging the gap between deep learning and field deployment, the proposed mobile system ensures low-latency execution and provides a rapid, reliable, and portable solution for improving operational safety in electrical fire investigations.

  • Open Access

    Android-Based Real-Time Classification of Electric Fire Short-Circuit Traces Using Lightweight Deep Learning Model

    by Mohammad Hadi Nazari, Junho Bang

    This paper presents a lightweight deep learning framework for classifying electric fire short-circuit traces to enhance safety and fault diagnosis in electrical energy systems. Accurate differentiation between primary (PSCT) and secondary short-circuit traces (SSCT) is essential for identifying failure origins, yet conventional manual inspection is time-consuming and subjective. To address these limitations, we systematically evaluate three lightweight convolutional neural network (CNN) architectures MobileNetV2, MobileNetV3, and EfficientNet using transfer learning on a domain-specific image dataset. The models are assessed based on accuracy, loss, precision, recall, and F1-score. Experimental results show that EfficientNet achieves the highest classification accuracy, while MobileNetV3 demonstrates the lowest validation loss and superior generalization stability. Based on a performance-efficiency trade-off analysis, MobileNetV3 is deployed on an Android platform using TensorFlow Lite, enabling real-time, offline, and on-device inference. To the best of our knowledge, this is among the first studies to integrate lightweight CNN-based short-circuit trace classification with real-time mobile deployment for on-site energy system fault analysis. By bridging the gap between deep learning and field deployment, the proposed mobile system ensures low-latency execution and provides a rapid, reliable, and portable solution for improving operational safety in electrical fire investigations.

  • Open Access

    Reinforcement Learning-Based Optimization of Environmental Control Systems in Battery Energy Storage Rooms

    by So-Yeon Park, Deun-Chan Kim, Jun-Ho Bang

    This study proposes a reinforcement learning (RL)-based optimization framework for the environmental control system of battery rooms in Energy Storage Systems (ESS). Conventional rule-based air-conditioning strategies are unable to adapt to real-time temperature and humidity fluctuations, often leading to excessive energy consumption or insufficient thermal protection. To overcome these limitations, both value-based (DQN, Double DQN, Dueling DQN) and policy-based (Policy Gradient, PPO, TRPO) RL algorithms are implemented and systematically compared. The algorithms are trained and evaluated using one year of real ESS operational data and corresponding meteorological data sampled at 15-min intervals. Performance is assessed in terms of convergence speed, learning stability, and cooling-energy consumption. The experimental results show that the DQN algorithm reduces time-averaged cooling power consumption by 46.5% compared to conventional rule-based control, while maintaining temperature, humidity, and dew-point constraint violation rates below 1% throughout the testing period. Among the policy-based methods, the Policy Gradient algorithm demonstrates competitive energy-saving performance but requires longer training time and exhibits higher reward variance. These findings confirm that RL-based control can effectively adapt to dynamic environmental conditions, thereby improving both energy efficiency and operational safety in ESS battery rooms. The proposed framework offers a practical and scalable solution for intelligent thermal management in ESS facilities.

  • Open Access

    An Intelligent Fault Detection Model for Fault Detection in Photovoltaic Systems

    by Barun Basnet, Hyunjun Chun, Junho Bang

    Effective fault diagnosis in a PV system requires understanding the behavior of the current/voltage (I/V) parameters in different environmental conditions. Especially during the winter season, I/V characters of certain faulty states in a PV system closely resemble that of a normal state. Therefore, a normal fault detection model can falsely predict a well-operating PV system as a faulty state and vice versa. In this paper, an intelligent fault diagnosis model is proposed for the fault detection and classification in PV systems. For the experimental verification, various fault state and normal state datasets are collected during the winter season under wide environmental conditions. The collected datasets are normalized and preprocessed using several data-mining techniques and then fed into a probabilistic neural network (PNN). The PNN model will be trained with the historical data to predict and classify faults when new data is fetched in it. The trained model showed better performance in prediction accuracy when compared with other classification methods in machine learning.