https://doi.org/10.5370/KIEE.2026.75.7.1451

김권철(Kwonchul Kim) ; 김인수(Insu Kim)

This study analyzes the structural characteristics of electricity demand at Incheon International Airport using long-term measured data from 2019 to 2024. The airport is pursuing a large-scale photovoltaic (PV) expansion plan to achieve RE100 by 2040, yet existing planning approaches are primarily based on peak demand and normal operating conditions. To address this limitation, this study focuses on quantifying the base load under both normal and abnormal conditions, including the COVID-19 period, which provides a natural experiment of extreme demand reduction. Base load is defined using the lower 10th percentile of seasonal daily electricity consumption to represent the minimum operational demand. The results show that electricity consumption did not decrease below a certain threshold even under extreme passenger decline, confirming the existence of a structurally sustained base load. Furthermore, a minimum base load level is observed even during abnormal conditions, indicating that conventional planning criteria may underestimate the lower bound of demand. These findings suggest that PV capacity planning should consider not only peak demand but also minimum load conditions to prevent reverse power flow. The identified base load characteristics can also be used to more accurately determine the required capacity of energy storage systems and the control range of energy management systems. This study provides practical insights for improving the reliability and robustness of renewable energy integration strategies in large-scale infrastructure systems.

https://doi.org/10.5370/KIEE.2026.75.7.1459

최연빈(Yeon-Bin Choi) ; 조형준(Hyeong-Jun Jo) ; 김우영(Woo-Young Kim) ; 김민철(Minchol Kim) ; 김수배(Soobae Kim)

This paper analyzes the impact of data center load models on the transient stability of power systems. Based on prior research, the study compares the Reference load model with the CLOD and ZIP load models. The ZIP load model tends to represent transient stability too optimistically when compared to the Reference load model, while the CLOD model tends to be slightly more conservative. The results show that the cooling load has the greatest impact on the CCT difference between the CLOD and Reference models, indicating that the dynamic characteristics of the cooling load have a significant effect on CCT. This paper emphasizes the importance of accurately modeling the cooling load in data center load models, highlighting that the ZIP model tends to overestimate transient stability, while the CLOD model provides a more balanced, realistic representation. The accurate modeling of cooling loads is essential for ensuring reliable data center operations and stability analysis in power systems.

https://doi.org/10.5370/KIEE.2026.75.7.1472

임세헌(Se-Heon Lim) ; 송경민(Kyung-Min Song) ; 류형우(Hyoung-Woo Ryu) ; 김대욱(Dae-Wook Kim) ; 윤성국(Sung-Guk Yoon)

This study quantitatively evaluates the economic trade-off between accepting solar power curtailment and deploying an Energy Storage System (ESS) for mitigation, using actual grid data from Jeju Island. Analysis of 2022? 2023 curtailment events shows that accepting curtailment is currently 31.03 million KRW per day more cost-effective than ESS deployment. Economic viability is primarily determined by the volume of curtailed energy and the volatility of the System Marginal Price (SMP): high SMP volatility supports ESS profitability via arbitrage, while low volatility favors accepting curtailment. The key tipping point for ESS economic advantage is a unit investment cost of 290 million KRW/MWh, projected for the late 2030s. Although ESS is critical for the long-term NDC target, these findings suggest accepting curtailment is currently economically justified and should be strategically used alongside future ESS expansion.

https://doi.org/10.5370/KIEE.2026.75.7.1480

황성욱(Sung-Wook Hwang) ; 박중성(Jung-Sung Park) ; 이휘성(Whee-Sung Lee) ; 김우석(Woo-Seok Kim) ; 권동영(Dong-Yeong Gwon) ; 최윤혁(Yun-Hyuk Choi)

As load growth and renewable energy integration continue to expand in distribution network, utilities face increasing challenges in reinforcing network infrastructure. Consequently, flexible resources such as ESS and DR have been introduced as NWAs. However, the economic contributions of these diverse resource configurations and their operational strategies have not been systematically evaluated for distribution network planning. Therefore, this paper proposes a framework for evaluating the economic feasibility between utilizing flexible resources such as ESS or DR and investing in distribution facilities, aiming to mitigate distribution network constraints. In particular, it is analyzed short-term and long-term operational strategies based on ESS, DR, and mixed DR/ESS configurations. This economic analysis framework can be used to support flexible resource allocation strategies and distribution facility investment plans in distribution network planning.

https://doi.org/10.5370/KIEE.2026.75.7.1489

안동혁(Dong-Hyuk Ahn) ; 김채린(Chae-Lyn Kim) ; 이도현(Do Hyeon Lee) ; 이병국(Byoung-Kuk Lee)

This paper proposes a control algorithm for achieving high-efficiency V2L operation over the entire load range in an 11-kW 2ph. Totem-pole PFC converter for electric vehicle on-board charger (OBC). To this end, the operating principles, key characteristics, and loss mechanisms of Full-bridge, Totem-pole, and interleaved Totem-pole control schemes are analyzed. Based on simulation-based comparative analyses of total harmonic distortion (THD) and power losses, the optimal control strategy is determined for each load condition, and a high-efficiency V2L control algorithm is developed accordingly. Finally, a hardware prototype of the 11 [kW] PFC converter is built, and experimental results under various load conditions are presented to validate the effectiveness of the proposed control algorithm.

https://doi.org/10.5370/KIEE.2026.75.7.1498

김현수(Hyeon-Soo Kim) ; 성선명(Seon-Myeong Sung) ; 이병국(Byoung-Kuk Lee)

This paper derives an optimal stacked working coil structure for high-speed heating in a semiconductor wafer induction heating (IH) system. The eddy current loss of the susceptor (Peddy) is proportional to the square of the switching frequency (fsw) and the maximum magnetic flux density (Bm). However, the equivalent parameters vary with the number of turns (N). A lower N results in lower equivalent inductance (Leq) and Bm, but requires higher fsw to meet the target equivalent resistance (Req). In contrast, increasing N increases Bm but reduces the fsw range satisfying Req. Therefore, the working coil design must consider the trade-off between fsw and Bm. Finite element analysis (FEA) based JMAG simulations are used to design single layer and double layer stacked coils and analyze heating characteristics. A 1kW half bridge-series resonant inverter (HB-SRI) based 12-inch class wafer IH system is implemented to experimentally verify the proposed coil selection.

https://doi.org/10.5370/KIEE.2026.75.7.1507

이상민(Sang-Min Lee) ; 조규진(Gyu-Jin Cho) ; 김현수(Hyeon-Soo Kim) ; 황윤성(Yun-Seong Hwang) ; 이병국(Byoung-Kuk Lee)

This paper proposes a high voltage gain DC-DC converter for fuel cell electric vehicle (FCEV) to enhance cost effectiveness. The proposed high voltage gain DC-DC converter can achieve a voltage gain on the order of ten, meeting the inverter input voltage demand for motor driving even under low fuel cell (FC) stack voltage conditions. Furthermore, the efficiency of the proposed DC-DC converter is measured over 96% under maximum power condition. The operating modes and principle according to duty ratio are theoretically analyzed, and its feasibility is verified through experimental results of a 2.6kW prototype.

https://doi.org/10.5370/KIEE.2026.75.7.1516

구재홍(Jae-Hong Koo) ; 오승열(Seungryle Oh) ; 정문규(Moon Gyu Jeong) ; 구교선(Kyo-Sun Koo) ; 박윤호(Yoonho Park) ; 이종건(Jong-Geon Lee)

This paper proposes a real-time dynamic operation system to ensure the equipment reliability of 362kV/6300A/63kA GIS circuit breakers under emergency overload conditions. These conditions result from the introduction of High Carbon Steel Core (HCSC) conductors, a measure taken to increase the capacity of 345kV transmission grids in response to the recent growth in renewable energy and power demand. The existing formula in IEC/TR 62271-306 recommend overload operation guidelines that take both temperature and current into account. This study presents a simplified formula that derives allowable duration based solely on current ratios by mathematically reducing complex temperature variable, alongside a three-stage serial priority determination logic that accounts for initial conditions. Furthermore, by benchmarking the practices of major international utilities, this study shifts away from fixed-table methods to establish a dynamic operation framework that tracks real-time ambient temperatures and load history, providing guidelines capable of responding effectively to both Short-Time Emergency (STE) and Long-Time Emergency (LTE) scenarios.

https://doi.org/10.5370/KIEE.2026.75.7.1524

기나혜(Nahye Ki) ; 양동제(Dongje Yang) ; 이선우(Sunwoo Lee) ; 김병훈(Byeonghoon Kim) ; 방수식(Su Sik Bang)

Accurate anomaly detection in power equipment depends on the selection of the time-series input window, which defines the temporal context available to learning models. Conventional practices rely on fixed window sizes based on hardware constraints or operational cycles, often disregarding the autocorrelation structure of multivariate sensor data. This study derives candidate window sizes from variable-specific critical lag times based on the autocorrelation function (ACF) and evaluates them on a transformer sensor dataset using LSTM, 1D-CNN, and Transformer encoder models. Experimental results show that several ACF-based candidates achieved higher macro F1-score than fixed daily and weekly baselines. In this dataset, Case 2 Method 2 was selected as the final criterion across the basic and variable-removal experiments, and Case 3 Method 1 showed that excluding non-valid variables improved average-based window estimation. These results indicate that ACF-based window-size selection provides a data-driven alternative to fixed temporal windows for multivariate time-series anomaly detection in power equipment.

https://doi.org/10.5370/KIEE.2026.75.7.1535

서태영(Taeyoung Seo) ; 이채원(Chae-Won Lee) ; 정경용(Kyungyong Chung)

Many existing intelligent CCTV systems rely on rule-based models, leading to frequent errors due to environmental factors, high resource burdens in large-scale environments, and inefficient manual configuration processes. Therefore, to address these issues, this study proposes an anomaly detection model that integrates deep learning-based object tracking technology with a hybrid clustering technique based on distance and density. The proposed method utilizes YOLO and DeepSORT algorithms to perform precise object detection and tracking, and the acquired spatiotemporal coordinate data undergoes a two-stage clustering process. First, the K-Means algorithm is employed to calculate the Cluster Centroid Distance to define primary movement paths and perform an initial screening of geometric deviations. Subsequently, the DBSCAN algorithm is applied to analyze spatiotemporal density, thereby finally detecting unstructured, rare anomalous trajectories. Experimental results confirm that the proposed model detects unstructured abnormal behaviors (such as S-shaped trajectories) more efficiently than simple distance-based analysis and significantly improves management convenience and detection accuracy through automated configuration.

https://doi.org/10.5370/KIEE.2026.75.7.1544

김민준(Min-jun Kim) ; 조현종(Hyun-chong Cho)

Accurate sow weight monitoring is essential for reproductive management and feeding strategies in commercial pig farms. Conventional weighing methods require direct handling, which is labor-intensive and stressful. To address this, we propose a 3D point cloud-based sow weight estimation framework that incorporates Center Point-Guided Oval Segmentation (CP-OS) for input refinement. Raw point clouds captured in farm environments often include irrelevant points such as floors and surrounding structures, distorting geometric features and degrading weight estimation accuracy. The proposed CP-OS isolates valid body regions and suppresses background noise, enabling more reliable geometric feature learning. We evaluated the framework using PointNet and Dynamic Graph CNN (DGCNN), and CP-OS improved performance for both backbones. Notably, PointNet combined with CP-OS achieved an average MAPE of 10.76% across three repeated trials, showing the best overall accuracy. These findings highlight the importance of geometric input refinement for robust 3D sow weight estimation in real farm environments.

https://doi.org/10.5370/KIEE.2026.75.7.1553

홍채희(Chae-Hui Hong) ; 유훈(Hoon Yoo)

This paper proposes an efficient and practical framework for thermal face recognition. The proposed method is designed as a Thermal-to-Visible (T2V) translator based on a residual-shifting diffusion model that converts thermal images into visible face images. The transformed images are then fed into a pre-trained high-performance visible-light face recognition model to extract and utilize facial embeddings. Thermal imagery contains limited visual cues for identity discrimination, and the lack of large-scale thermal training data restricts performance improvements through direct model retraining. To address this issue, the proposed method integrates modality-specific autoencoders with a residual-shifting diffusion process to develop a T2V translator optimized for modality transformation. Compared with conventional diffusion-based generative models, the proposed method significantly improves inference efficiency by performing the transformation with only 15 sampling steps. The generated visible images are input into a standard visible-light face recognition network to extract facial embeddings, which are subsequently used for final identity recognition. Experimental results on the SpeakingFaces dataset demonstrate that the proposed approach significantly improves both face detection and recognition performance when applied to a visible-light recognition pipeline, confirming that the proposed framework achieves both high inference efficiency and effective thermal face recognition.

https://doi.org/10.5370/KIEE.2026.75.7.1562

김민지(Minji Kim) ; 정경용(Kyungyong Chung)

In this study, we propose a wav2vec 2.0-based model for early dementia detection from Korean daily speech, validated on the AI Hub dataset of 5,769 recordings from 1,002 elderly speakers labeled as normal cognition, mild cognitive impairment (MCI), or Alzheimer's disease (AD). The model integrates a multilingual SSL (XLSR-53) and Korean ASR-adapted wav2vec 2.0 backbone, CNN feature encoder, 24-layer Transformer, and attentive statistics pooling head. Fine-tuned with Focal Loss, it effectively handles spontaneous speech, environmental noise, and class imbalance. Five-fold cross-validation achieves speaker-level (10 8-second segments, based on soft voting) 94.5% accuracy, 92.9% recall, and ROC-AUC of 0.958, outperforming handcrafted-feature baselines, Audio Spectrogram Transformer, and English-pretrained HuBERT. Future work will explore knowledge distillation for on-device inference, multimodal fusion with linguistic features, and explainable AI for real-time biomarker assessment.

https://doi.org/10.5370/KIEE.2026.75.7.1571

이상정(Sang-Jeong Lee) ; 서성발(Sung-Bal Seo) ; 배유석(You-Suk Bae)

We propose a practical VLM?RAG pipeline for PCB defect classification under severe label scarcity. Document-grounded prompts (from inspection specs and SOPs) align a CLIP-style zero-shot classifier to the manufacturing domain, ROI-tiling amplifies localized cues at inference, and a small-sample linear probe on frozen embeddings sharpens decision boundaries with minimal labeling. On a real PCB dataset, we compare domain-tuned zero-shot, ROI-tiling zero-shot, and the linear probe. ROI-tiling yields a small accuracy gain over zero-shot but does not improve macro-F1, while the linear probe achieves a meaningful improvement in both accuracy and macro-F1 over the zero-shot baseline. The reported accuracies remain below the level required for direct deployment; however, the results indicate that combining Visual RAG, document-grounded prompts, and lightweight supervision can reduce the cost of building an initial defect-classification model in data-scarce manufacturing environments, providing a practical starting point that can be incrementally refined as more labels become available.

https://doi.org/10.5370/KIEE.2026.75.7.1579

이찬석(Chan-Seok Lee)

This paper proposes an IMM-based vertical-plane homing filter for bearing-only homing missiles engaging low-altitude surface targets with uncertain altitude. In a conventional augmented extended Kalman filter (A-EKF), target altitude is estimated with X-axis position from a single line-of-sight (LOS) angle measurement. This coupling makes the Fisher information matrix ill-conditioned under low LOS-rate conditions, degrading estimation accuracy and possibly causing divergence. The proposed filter partitions target altitude uncertainty into a finite set of hypotheses and employs altitude-conditioned subfilters. Treating altitude as a known parameter in each subfilter removes the X-axis position-altitude coupling and improves X-axis state observability. The IMM mode probabilities are interpreted as Bayesian posterior probabilities for altitude estimation. Monte Carlo simulations show improved LOS-rate estimation accuracy and reduced miss distance over conventional methods.

https://doi.org/10.5370/KIEE.2026.75.7.1591

한동훈(Donghun Han) ; 오우진(Woojin Oh) ; 윤희철(Heechul Yoon)

Acoustic focusing is one of the fundamental components of ultrasound systems for both diagnostic and therapeutic purposes. Accordingly, each pulser is dedicated to controlling the active elements of the ultrasound transducer with different time delays to achieve precise focal positioning. However, mapping each element to a dedicated pulser linearly increases hardware complexity, which poses significant challenges in the recent adoption of large-scale element arrays such as two-dimensional arrays. In this paper, we propose a pulser quantization method that exploits the symmetricity of focusing delays in two-dimensional array?based systems while accounting for the finite sampling period of the pulser clock. The impact of pulser quantization on focusing performance is systematically analyzed, and an effective approach to reduce the number of required pulsers while preserving focusing capability is presented. By incorporating allowable focusing error for localized therapeutic applications, the proposed method achieves pulser count reductions of approximately 50%, and up to 87% when therapeutic focusing tolerance is considered, suggesting applicability to two-dimensional array?based 3D ultrasound systems and portable ultrasound devices.

https://doi.org/10.5370/KIEE.2026.75.7.1601

민지희(Jihee Min) ; 김주현(Juhyun Kim) ; 이언석(Onseok Lee)

Nasal obstruction reduces nasal airflow and induces mouth breathing. Conventional tests (anterior rhinoscopy, computed tomography) are limited by examiner dependence, radiation, and capture of flow dynamics. We quantified nasal expiratory airflow from thermal videos of 35 adults. After Otsu segmentation, the nostrils were separated and the lowest endpoint of the exhaled thermal plume was tracked. Trajectories were fitted with parametric polynomials using a normalized parameter u∈[0,1]; degree (1-10) was selected by Ridge regression with 5-fold CV. We extracted morphological and physiological features for the analysis. Features were clustered by k-means (elbow: k=2). Max acceleration, mean curvature, degree, and extrema count differed, with Cluster 0 showing more complex, rapidly varying flows overall. Through this analysis, we aimed to propose quantitative indicators that can be used to assess nasal function. These findings is expected to contribute to clinical screening and monitoring.

https://doi.org/10.5370/KIEE.2026.75.7.1611

권오극(Ogeuk Kwon) ; 이광호(Gwang-Ho LEE) ; 오힘찬(Him-Chan Oh) ; 차한주(Hanju Cha)

Surge arresters play a critical role in protecting transmission lines and power generation equipment by discharging temporary overvoltages to the ground. Since the mid-1980s, Zinc Oxide (ZnO) arresters have been widely utilized in power plants, substations, and transmission lines. The operating voltage is applied directly to the ZnO blocks, resulting in a continuous flow of a minute leakage current. ?This steady-state leakage current is a composite of capacitive and resistive components. While an increase in the resistive leakage current indicates arrester degradation, accurately measuring this component in the field remains a significant challenge. Since 2000, various techniques have been reported to estimate resistive current, such as analyzing third-harmonic components or extracting the resistive portion from the total leakage current by applying an external voltage. ?In this study, a resistive leakage current detection algorithm based on phase analysis was implemented to measure total leakage current without voltage application, utilizing A/D conversion and microprocessors. A specialized measurement device was designed and fabricated, and its performance was validated through comparative tests with voltage-injection type equipment in a laboratory setting. Furthermore, field measurements were conducted overseas, confirming the thermal degradation of arresters where leakage current levels exceeded standard thresholds.

https://doi.org/10.5370/KIEE.2026.75.7.1619

구자영(Ja-Young Koo) ; 곽준호(Jun-Ho Kwak) ; 김희동(Hee-Dong Kim)

As the stator windings of weather protected type II(WPII) high-voltage(HV) motors undergo prolonged operation, degradation occurs through interfacial delamination and void formation within the groundwall insulation(GWI). Additionally, the outer corona protection (OCP) conductive layers deteriorate within the slot and slot-exit regions, while environmental contaminants accumulate on the core and endwinding surfaces. These factors not only amplify partial discharge(PD) magnitudes but also induce significant shifts in phase-resolved PD(PRPD) patterns. This paper presents insulation condition assessments performed on five 11 kV WPII motors operated for seven years at a 1,000 MW coal-fired power plant. For motors No. 1 and No. 2(2,700 kW), the AC current increment (△I), dissipation factor increment(Δtanδ), and PD magnitudes remained within nominal thresholds at the rated phase voltage(6.35 kV). However, critical PD activity was detected at 1.25 times the phase voltage(7.94 kV), necessitating a proactive maintenance recommendation. Furthermore, for motors No. 3 through No. 5(1,500 kW), a longitudinal analysis of PD characteristics was conducted after a two-year post-maintenance interval to verify the long-term efficacy and dielectric stability of the applied insulation reinforcement.

https://doi.org/10.5370/KIEE.2026.75.7.1626

츄방제(Bangjie Qiu) ; 류호영(Ho-Young Ryu) ; 오윤식(Yun-Sik Oh)

Accurate mapping of local IDs across heterogeneous power system applications is essential for data integration but remains challenging due to inconsistent naming and data models. This paper proposes a topology-based graph convolutional neural network (GCN) method for automatic local ID mapping of power equipment. The power network is modeled as a graph, and six topological features are used to represent node characteristics. A multi-layer GCN learns topology-aware node embeddings, and the Hungarian algorithm is applied to obtain optimal one-to-one correspondences based on embedding similarity. Simulation results on a 1500-node test network with loops representing symmetric structures demonstrate that the proposed method achieves over 95% accuracy and edge consistency with the help of GCN trainings. In addition, margin analysis confirms improved separability between correct and incorrect mappings, indicating high robustness. The proposed approach provides an effective and scalable solution for topology-aware ID mapping in power systems.

https://doi.org/10.5370/KIEE.2026.75.7.1633

안정열(Jeong-Yeol An)

Generally, conventional Brushless DC (BLDC) motor drives typically rely on physical transducers, such as encoders or Hall-effect sensors, to acquire rotor position feedback. To enhance system robustness and reduce costs, sensorless control for trapezoidal-type BLDC motors has become a critical area of research. This paper introduces an innovative commutation signal extraction technique that derives rotor position directly from the Back Electromotive Force(Back-emf) and using a multi-function OP-AMP circuit. However traditional sensorless method that necessitate motor neutral voltage monitoring, multi-stage analog filtration, or computationally intensive digital phase-lag compensation, the proposed architecture streamlines the signal processing. By synthesizing commutation triggers from the integrated average phase voltage rather than identifying the zero-crossing points (ZCP) relative to a floating neutral, the system eliminates the requirement for a physical or virtual neutral point. This approach significantly bolsters immunity against common-mode switching noise and simplifies the hardware-software interface. The effectiveness and precision of the proposed control algorithm is verified through simulation results.

https://doi.org/10.5370/KIEE.2026.75.7.1639

김상호(Sang-Ho Kim) ; 김일송(Il-Song Kim)

This paper proposes a precise analysis and modeling technique for a coupled inductor based boost converter with a symmetrical winding and identical inductance structure. Previous studies have tended to simplify the analysis by restricting the coupling coefficient() to a specific range or by using current conversion methods based on the turns ratio. Consequently, there have been limitations in precisely examining the impact of the coupling coefficient() and the actual secondary current on system dynamics and control stability. To address these issues, this study performs precise modeling that simultaneously considers the coupling coefficient() and secondary current by reflecting the practical electrical characteristics of the coupled inductor.[1-5]. The proposed method derives state-space equations for each operating mode based on voltage and current differential equations and transforms them into a small-signal model using the state-space averaging method. In particular, by applying the identical inductance condition(), the design variables are optimized to the inductance() and coupling coefficient(), thereby reducing analytical complexity and enhancing practical utility. The validity of the proposed model was verified through double-loop controller design using MATLAB SISOTOOL and PSIM simulations. The results demonstrated a high correlation between the theoretical step response and the switching non-linear simulation waveforms, proving the precision of the proposed design parameters and the effectiveness of the system stability control.

https://doi.org/10.5370/KIEE.2026.75.7.1648

안창환(Chang-Hwan Ahn) ; 이강희(Kang-Hee Lee)

This paper analyzes the power factor characteristics of apartment building electrical systems under harmonic conditions based on field measurement data. Long-term continuous measurements of voltage, current, power factor, and current harmonics were conducted at the transformer primary and secondary sides, low-voltage distribution panels, and circuit breaker levels to analyze time-dependent characteristics. The results show that excessively leading power factor conditions were continuously observed over a 24-hour period at most measurement points, while current harmonic distortion levels exceeding international standards were simultaneously measured. In particular, as harmonic components increased, a discrepancy between the displacement power factor and the true power factor occurred, indicating that conventional power factor indices based on displacement power factor alone are insufficient to adequately explain the power quality characteristics of apartment building electrical systems. These results indicate that the electrical systems of apartment buildings are facing very high levels of risk, and that the establishment of comprehensive power factor evaluation methods and rational management criteria that consider harmonic effects is urgently required.

https://doi.org/10.5370/KIEE.2026.75.7.1659

이재윤(Jae-Yoon Lee) ; 황영익(Young-ik Hwang) ; 윤형익(Hyoung-Ik Youn) ; 이연수(Yeon-Su Lee) ; 이현우(Hyeon-Woo Lee) ; 이종순(Hyun-Tae Kim) ; 김현태(Jong-Soon Lee) ; 김규호(Kyu-Ho Kim)

This study aims to redefine the value and management approach of private customers’ electrical equipment assets by enabling safer and more efficient maintenance and replacement practices. To address the limitations of conventional methods in detecting early signs of equipment abnormalities, we introduce real-time diagnostic technologies that enhance the reliability and responsiveness of condition assessment. The proposed approach integrates IoT-based sensing and AI-driven anomaly detection to improve the accuracy of equipment monitoring and support continuous, non-interruptive operation. Furthermore, based on these technologies, we seek to develop a big-data standard platform that can contribute to broader public safety by enabling systematic data collection, predictive maintenance, and efficient safety management. Ultimately, the outcomes of this research are expected to strengthen preventive maintenance capabilities and enhance the overall safety and operational reliability of electrical equipment.