https://doi.org/10.5370/KIEE.2026.75.3.443

이도형(Do-Hyeong Lee) ; 김세호(Se-Ho Kim)

This paper forecasts the future share of renewable energy generation in the Jeju power system by jointly considering the expansion of renewable energy facilities and the growth of electricity demand. To this end, multiple electricity demand models are developed based on the Basic Plan for Electricity Supply and Demand and are combined with renewable energy capacity plans to analyze annual and time-of-day generation shares. In addition, the variation in the share of renewable energy generation is compared under different operating conditions of HVDC with the mainland power system. The results show that, although renewable energy generation increases, its share in total electricity supply exhibits different trends depending on changes in electricity demand, and that HVDC operating conditions have a significant impact on renewable energy hosting capacity. Based on these findings, this paper provides insights that can be used to support future power system operation and energy policy formulation aimed at expanding renewable energy.

https://doi.org/10.5370/KIEE.2026.75.3.450

유민지(Min-Ji Ryu) ; 허진(Jin Hur)

As the penetration of renewable energy increases, the use of high-resolution time-series data becomes indispensable in long-term power mix planning. This study employs a chronology-based fitting approach to evaluate the accuracy and computational efficiency of reduced time-series in Jeju’s power mix simulations. To address the inherent limitations of conventional reduction techniques, an Integrated Chronology Framework is proposed that explicitly reflects the distinct temporal characteristics of load, solar, and wind resources. The proposed framework preserves critical chronological patterns while ensuring computational tractability, which is essential in systems with high renewable variability. Future research will focus on conducting comprehensive sensitivity analyses across diverse operational scenarios and extending the methodology to mainland data, thereby strengthening the robustness and applicability of long-term power mix planning in support of sustainable energy transition.

https://doi.org/10.5370/KIEE.2026.75.3.460

이효(Hyo Lee) ; 이동윤(Dong-Yoon Lee) ; 김남규(NamKyu Kim) ; 정주용(JooYong Jung) ; 양재영(JaeYoung Yang) ; 권영진(YoungJin Kwon) ; 심재웅(Jae Woong Shim)

This paper introduces a novel farm-level controller designed to ensure stable and compliant grid integration for large-scale renewable energy power plants under Korea Electric Power Corporation (KEPCO) Technical Connection Code. In renewable farms operating without a plant controller, the uncoordinated control of individual generators, combined with internal cable impedances, creates challenges in managing voltage and reactive power at the Point of Measurement (POM), thus hindering KEPCO compliance. To address this, our proposed controller directly monitors POM conditions and coordinates turbine output through three functions: dynamic operation mode determination, precise active/reactive power setpoint calculation, and intelligent weighting factor distribution. PSCAD/EMTDC simulations of offshore wind farms confirm its effectiveness. The controller successfully met all KEPCO reactive power standards, including supply capability, output control, power factor, and voltage-reactive power control, which individual generator control could not achieve. Furthermore, fault ride-through simulations, utilizing a proposed current limitation, demonstrated robust grid support via preferential reactive current injection for voltage recovery. A STATCOM capacity assessment was also conducted. Ultimately, this farm-level controller significantly enhances reactive power control, vital for grid stabilization, and validates its adherence to KEPCO's stringent technical criteria.

https://doi.org/10.5370/KIEE.2026.75.3.472

최태완(Taewan Choi) ; 최윤성(Yoon-Seong Choi) ; 임철훈(CheolHun Im) ; 소순열(Soon-Yeol So) ; 이동호(Dongho Lee)

The increasing penetration of renewable energy and power electronic devices has weakened grid strength by reducing system inertia and short-circuit capacity. In such weak-grid conditions, converter control loops interact with network impedance, causing oscillations across a wide frequency range from low-frequency to sub- and super-synchronous modes. This paper proposes an impedance-based reflection coefficient method for the probabilistic pre-detection of oscillation-vulnerable buses before system disturbances occur. The method computes from the system and load impedances at each bus and analyzes variations in its magnitude and phase with operating conditions (P,Q) Based on the Barkhausen stability criterion, buses showing abrupt phase changes are identified as oscillation-vulnerable points within the network. Simulation studies on the IEEE 39-bus using PSS/E verify the effectiveness of the proposed probabilistic pre-detection approach.

https://doi.org/10.5370/KIEE.2026.75.3.481

조재경(Jae-gyeong Jo) ; 백건(Keon Baek)

This paper conducted a correlation analysis to determine the impact of renewable energy generation on system marginal prices in the CBP and PBP markets, and compared the results. Data on renewable energy generation, system marginal prices, and electricity demand at hourly intervals were used for Korea and California. The analysis revealed a strong negative correlation in the PBP market, but a minimal correlation in the CBP market. This correlation stems from differences in market structure and the proportion of renewable energy generation facilities deployed. Furthermore, the impact is expected to intensify in Korea as renewable energy deployment expands. By quantitatively comparing the relationship between renewable energy and system marginal prices, this study suggests directions for future research.

https://doi.org/10.5370/KIEE.2026.75.3.489

이세미(Semi Lee) ; 정재성(Jaesung Jung)

Increasing penetration of distributed energy resources (DER), particularly electric vehicles (EV), challenges distribution system operations with issues such as voltage instability and congestion. Effectively managing the flexibility from numerous, uncertain EVs requires advanced frameworks. This paper proposes a novel hierarchical control strategy for maximizing EV flexibility in a distribution system. The strategy features multiple layers enabling efficient upstream aggregation and downstream control allocation. Individual EV flexibility, quantified with opportunity cost, is aggregated and simplified using curve fitting. The top-level entity perform network-aware optimization considering grid constraints to determine flexibility request for lower-level entities. Subsequently, lower-level entities execute cost-effective scheduling for their EV portfolio within these targets and local constraints. Simulations on the IEEE 33-bus system validate the approach.

https://doi.org/10.5370/KIEE.2026.75.3.500

이태승(Tae-Seung Lee) ; 송유훈(Yu-Hoon Song) ; 최지원(Jiwon Choi) ; 국경수(Kyung Soo Kook)

This study presents a quantitative evaluation method for determining the optimal location and required capacity of Battery Energy Storage Systems (BESS) which is expected to be continuously installed in the power systems with a high penetration level of the renewable energy resources. For doing this, a new index, the Frequency Control Effectiveness Index(FCEI), is proposed to assess the frequency control performance of the power system by comparing the frequency distortion magnitude with the performance criteria, providing a normalized measure of the control performance. Contribution of BESS on the frequency control can be evaluated by calculating FCEI according to its installation location and system operating conditions. Additionally, a system-level index (SYS_FCEI) is introduced to comprehensively quantify the frequency control performance of various operating conditions of the power systems. By using SYS_FCEI, the optimal location and required capacity for installing BESS can be determined considering the distribution of installation locations. The proposed method was validated through intensive study cases employing Korean power system database developed in accordance with the Basic Plan for Electricity Supply and Demand.

https://doi.org/10.5370/KIEE.2026.75.3.509

이동주(Dong-Ju Lee) ; 김종겸(Jong-Gyeum Kim)

For small-scale pumped storage power plants, induction machines are more commonly used than synchronous machines in terms of initial investment cost and maintenance. Induction machines require reactive power for magnetization in motoring or generating modes. In particular, even if the output is the same, the reactive power is different in motoring or generating operation, so there is a difference in power factor. Power companies require that the power factor of induction machines connected to the grid always be maintained above a certain value. This study compared and simulated the changes in power and power factor when an induction machine with the same output was used in motoring mode and generating mode without a power factor correction device, when a compensation device such as a capacitor was used, and when an active compensation device was installed. The simulation results confirmed that compensation based on reactive power demand in generating mode effectively meets power factor requirements in both motoring and generating mode. Furthermore, the use of STATCOM maintained the power factor close to 1 regardless of operating mode or load fluctuations. This study result will contribute to improving the power factor and power quality of induction machines in small-scale pumped storage systems requiring both motoring and generating operation.

https://doi.org/10.5370/KIEE.2026.75.3.515

박종훈(Jong-Hoon Park) ; 한지훈(Ji-Hoon Han) ; 홍선기(Sun-Ki Hong)

The hysteresis motor had a characteristic in which the rotor ring volume is proportional to the output power. Therefore, increasing the ring thickness can enhance output by enlarging the rotor volume. However, beyond a certain thickness, the output no longer increases proportionally and even decreases due to rotational hysteresis effects. Previous designs based on equivalent circuits could not fully consider this phenomenon. In this study, finite element analysis capable of reflecting rotational hysteresis was used to verify this effect analytically. By varying the axial length and ring thickness under the same rotor volume, the range of thicknesses causing output reduction was identified. Furthermore, this behavior was analyzed more reliably through calculations of the phase difference between magnetization and magnetic field. The findings are expected to contribute to improving the output performance of small hysteresis motors.

https://doi.org/10.5370/KIEE.2026.75.3.522

박현정(Hyeon-Jeong Bak) ; 김해성(Hae-Seong Kim) ; 김병오(Byung-Oh Kim) ; 홍영희(Young-Hee Hong) ; 서동관(Dong-Kwan Seo) ; 정승욱(Seung-Wook Jung)

This study presents the development of a domestically designed reactor coolant pump (RCP) motor for the OPR1000 nuclear power plants. The motor is designed based on electromagnetic analysis to satisfy the required performance despite the reinforced insulation. Furthermore, rotor dynamics analysis is performed to verify the mechanical reliability of the rotor assembly. The analysis results are validated through manufacturing a prototype and experimental testing.

https://doi.org/10.5370/KIEE.2026.75.3.532

심민주(Min-Ju Sim) ; 한동호(Dong-Ho Han) ; 김종훈(Jong-Hoon Kim)

Lithium-ion batteries (LIBs) play a pivotal role in energy storage systems (ESSs) and electric vehicles (EVs); however, they inevitably undergo capacity fading and performance degradation during long-term operation. Accordingly, accurate state-of-health (SOH) estimation is essential for reliable battery management. In this study, refined health indicators (HIs) are defined from charge?discharge voltage?time characteristics, and a flexible SOH prediction framework is proposed. Aging data obtained from INR21700-33J cells over 1,400 cycles are analyzed to extract multiple HIs, including mean voltage falloff (MVF), voltage interval of equal discharging/charging time difference (VIEDTD/VIECTD), and time interval of discharging/charging voltage difference (TIEDVD/TIECVD). These indicators are further subdivided using voltage resolutions of 0.1 V and 0.01 V, enabling robust HI extraction under data imbalance. To compensate for missing HI data caused by external operational factors in practical ESS environments, a denoising autoencoder (DAE)-based interpolation method is employed to preserve temporal degradation trends. Subsequently, recurrent neural network (RNN), long short-term memory (LSTM), and gated recurrent unit (GRU) models are applied for SOH prediction. Among them, the GRU model achieves the best performance, with an MAE of 2.24, RMSE of 2.74, and R² of 0.97, demonstrating improved accuracy and robustness under incomplete operational data conditions.

https://doi.org/10.5370/KIEE.2026.75.3.542

이희찬(Heechan Lee) ; 이동철(Dongcheol Lee) ; 김종훈(Jonghoon Kim)

This study presents a data-driven framework for early detection of abnormal cells in lithium-ion battery modules. Incremental capacity analysis (ICA) is performed on current?voltage data to visualize cell degradation as ICA curves, which are then used as inputs to Autoencoder(AE) models for training and detection. Two 6S2P lithium-ion battery modules were tested: one with overcharged cells (positions 1 and 6) and the other with overdischarged cells (positions 1 and 6). Each cell underwent 100 charge?discharge cycles under constant current (CC) conditions to simulate progressive aging. From these, 80 ICA cycles from normal cells (positions 2?5) were used for training, and the rest for evaluation, ensuring the model only learned healthy behavior patterns. Four detection approaches were compared: ICA vectors, ICA images, Recurrence plot (RP) images, and RGB-based multi-channel time series encoding using Gramian angular field(GAF) and Markov transition field(MTF) images. Models were trained only on normal cell data, and anomalies were identified by increases in reconstruction error, such as Mean squared error (MSE). Results showed that, in terms of mean F1-score, RGB GAF?MTF delivered the best performance (0.841), followed by RP (0.775), whereas ICA-vector AE and ICA-image CNN lagged behind (0.673 and 0.504, respectively), underscoring the advantage of multi-channel time-series encodings for capturing subtle degradation.

https://doi.org/10.5370/KIEE.2026.75.3.552

김현석(Hyun-Seok Kim) ; 이상민(Sang-Min Lee) ; 김현수(Hyeon-Soo Kim) ; 황윤성(Yun-Seong Hwang) ; 이병국(Byoung-Kuk Lee)

In fuel cell electric vehicles (FCEVs), the fuel cell DC?DC converter (FDC) boosts the voltage of the FC stack to the DC-link voltage required by the inverter. To reduce the volume of the FC source by lowering the FC stack voltage, a topology that can achieve higher voltage gain than conventional boost converter is required. Therefore, a cascaded boost converter (CBC) which has a large voltage gain can be adopted. Given that the CBC can apply various control method, a quantitative analysis under each control method is required to determine high-efficiency operation. Therefore, this paper analyzes and compares three control methods for CBC to determine the optimal control strategies considering FC voltage and current. A 55kW multi?phase interleaved CBC is designed considering volume, efficiency, and temperature of power devices. Furthermore, theoretical loss and efficiency analyses are performed and validate the proposed control strategy’s effectiveness through experiments results.

https://doi.org/10.5370/KIEE.2026.75.3.562

이동주(Dong-Ju Lee) ; 김종겸(Jong-Gyeum Kim)

This paper proposes a hybrid power system(HPS) that integrates Pumped-Storage Hydroelectricity(PSH) with Floating Photovoltaic (PV) power generation, utilizing the underused infrastructure of a agriculture reservoirs as a mechanical Energy Storage System(ESS). We quantitatively applied this HPS model to an operational reservoir to evaluate its technical and economic viability. The analysis, conducted using Electro-Magnetic Transients Program (EMTP) simulations, demonstrates that the system effectively maintains a stable grid frequency of 60 Hz by mitigating the intermittency of PV power through real-time MPPT control and pumped-storage synergy. Initial economic assessments yielded a Benefit-Cost Ratio (BCR) of 0.68, indicating low feasibility for the baseline model. However, by expanding the floating PV capacity fourfold to leverage economies of scale and maximize the utilization of the available water surface, the BCR improved to 1.1. This enhancement is attributed to the distribution of fixed costs over a larger generation capacity and the elimination of expensive chemical battery requirements, as the water body itself serves as the storage medium. These findings suggest that the proposed HPS model can serve as an efficient decentralized power solution to support carbon neutrality goals.

https://doi.org/10.5370/KIEE.2026.75.3.569

신승협(Seung-Hyeop Shin) ; 김윤영(Yoon-Yeong Kim)

This study proposes a novel Debiased Active Learning (DAL) approach for solar photovoltaic (PV) power prediction based on ground-based sky images. Conventional uncertainty-based sampling methods often suffer from selection bias, particularly under class-imbalanced conditions where clear-sky samples dominate over cloudy-sky samples. To address this issue, the proposed DAL method estimates a debiasing matrix from a small, trusted validation dataset and uses it to correct the model’s predictive probabilities before uncertainty sampling. Using the SKIPP’D (Sky Images and Photovoltaic Power Dataset) dataset(?300,000 paired sky images and PV power values), the proposed DAL significantly improves labeling efficiency and prediction accuracy compared to traditional uncertainty-based active learning approaches.

https://doi.org/10.5370/KIEE.2026.75.3.577

김가현(Ga-Hyun Kim) ; 서예슬(Ye-Seul Seo) ; 박건희(Keon-Hee Park) ; 임장섭(Jang-Seob Lim)

Aluminum conductors are used in overhead transmission lines, while copper is used in underground and submarine cables. This study analyzed thermal characteristics of compression, bolted, and welded sleeves for aluminum, and compression and bolted sleeves for copper. Conductive compound and moisture compound removal effects on sleeve heating were evaluated under varying conditions. Thermal cycling tests per ANSI C119.4 and C119.0 standards (60?90℃) measured temperature differences between sleeves and reference conductors. Statistical analysis using Weibull and normal distributions showed bolted connection had best performance. Conductive compound applied with two outer moisture layers removed showed the lowest temperature. Similar patterns were observed at 600 A for aluminum and 900 A for copper, confirming aluminum's lower conductivity affects heating. Results provide baseline data for evaluating 70 kV aluminum underground and submarine cable joints..

https://doi.org/10.5370/KIEE.2026.75.3.583

장종인(Jong In Jang) ; 정해창(Hae-Chang Jeong)

This study focuses on the development and analysis of polyvinylidene fluoride (PVDF) alignment layers enhanced with silver nanowires (SNWs) to improve the electro-optic properties of liquid crystal (LC) devices. Optical characterization using UV-Vis spectroscopy showed that all SNW-PVDF films maintained high transmittance (~85?86%), suitable for display applications. Atomic force microscopy (AFM) confirmed stable surface morphology with no significant disruptions from SNW inclusion. Electro-optic analysis revealed a reduction in threshold voltage and improved response times, particularly for films with lower SNW concentrations. Finite element method (FEM) simulations highlighted the mechanism by which SNWs enhance local electric fields (~1.5×) near alignment layers, facilitating faster LC driving. Additionally, SNWs mitigated hysteresis effects observed in the voltage-capacitance curves, contributing to device stability. These findings demonstrate that SNW-PVDF alignment layers provide an effective pathway for enhancing LC device performance, offering advancements for electro-optic applications.

https://doi.org/10.5370/KIEE.2026.75.3.590

사공준(Jun Sagong) ; 한희제(Heeje Han) ; 김동원(Dong-won Kim) ; 박진광(Jin-gwang Park) ; 김홍준(Hongjoon Kim)

Recently, the use of higher frequencies in next-generation wireless communication environments, such as 5G and 6G, has significantly improved the speed of data transfer. However, in a dense urban area, high-frequency signals suffer a severe shadowing and fading due to many obstacles. In this paper, we propose a reflective electromagnetic-wave-splitting metasurface to solve these problems. The designed metasurface operates at 5.8 GHz which is a frequency commonly used in the ISM band for urban applications, splitting incident waves under normal incidence into ±58° while suppressing specular reflection significantly, and achieves a wider beam-splitting angle compared to the previous studies. By reducing the disruptions caused by obstacles, the proposed reflective metasurface is expected to improve the signal fidelity of the wireless communication system and mitigate the signal disruption.

https://doi.org/10.5370/KIEE.2026.75.3.595

(Hanyang Li) ; 권태욱(Taewook Kwon) ; 홍재연(Jaeyeon Hong) ; 박세은(Se-eun Park) ; 나완수(Wansoo Nah)

This paper analyzes electromagnetic radiation in the 2.4 GHz band generated during high-speed data transmission via USB Type-C connectors and proposes a shielding solution to mitigate near-field emissions. Motivated by the EU's common-charger policy and the rapid adoption of USB-C, which heighten EMI risks for Bluetooth and other services in this band. We quantitatively measured both near-field and far-field radiation under active data-link conditions. Near-field scans revealed spurious emissions around 2.4 GHz at levels that may exceed regulatory criteria or cause wireless interference. Guided by these measurements, we modeled and designed an effective shielding structure using electromagnetic simulation software Ansys HFSS Simulation results indicate a substantial reduction of radiation in the problematic band, suggesting a practical near-field shielding approach for USB Type-C data-transmission modes.

https://doi.org/10.5370/KIEE.2026.75.3.605

정민경(Minkyung Jeong) ; 김예찬(Yechan Kim) ; 구본혁(Bonhyuk Ku) ; 김재상(Jaesang Kim) ; 강형구(Hyoungku Kang)

Power networks are long-lived, capital-intensive assets facing complex risks. PAS 55 and ISO 55000 provide general asset-management baselines but lack power-system specificity. The IEC TC 123 IEC 63223 series addresses this gap with a domain-tailored framework. This paper reviews its three-layer architecture (Core?Process?Support), contrasts it with PAS 55/ISO 55000, and assesses implementation readiness. IEC 63223-1 sets strategic principles and four decision domains: asset development, operation, resource portfolio, and life-cycle strategy. IEC 63223-2 formalizes risk-informed decision making (RIDM) using PoF, CI, and HI, and classifies uncertainty as aleatory, epistemic, or normative. IEC TS 63224 contextualizes ISO 55001 and provides actionable guidance (KPIs, SAMPs). Findings show greater adaptability, sector specificity, and alignment with digitalization, ESG, and PHM, offering practical implications for operators, regulators, and OEMs, and a concise roadmap for adoption.

https://doi.org/10.5370/KIEE.2026.75.3.615

이훈희(Hoon Hee Lee) ; 김도완(Dowan Kim)

This paper addresses the path-following problem of an unmanned surface vehicle (USV) that must track a predefined path while maintaining a constant surge speed. To this end, the surge speed and yaw rate errors of the USV are formulated as sliding variables and a barrier function-based adaptive sliding mode controller is designed to ensure that the surge speed and yaw rate error converge within a predefined barrier. By incorporating a quasi?positive definite barrier function into the adaptation law, the proposed method mitigates the gain overestimation and the resulting chattering issues commonly observed in conventional adaptive sliding mode control. Unlike existing adaptive sliding mode control schemes, the proposed control strategy allows the convergence region and convergence time to be explicitly defined independently of the upper bound of disturbances, thereby preserving the sliding motion even when disturbances vary rapidly. Furthermore, the Lyapunov stability theory is employed to theoretically establish the finite-time stability of the closed-loop system. In addition, numerical simulations are conducted to verify the performance of the proposed controller.

https://doi.org/10.5370/KIEE.2026.75.3.624

채나정(Najeong Chae) ; 최지호(Jiho Choi) ; 변성우(Sung-Woo Byun) ; 이혜민(Hea-Min Lee)

Numerous studies have been conducted on smart agriculture, which integrates information and communication technologies across the entire agricultural process. Smart agriculture incorporating various technologies enhances productivity, reduces labor requirements, and enables the production of high-quality crops. LiDAR, a technology that measures object position coordinates and distances, offers a wide field of view and high-precision measurements, and has been applied to smart agriculture in combination with other information and communication technologies. In addition, advanced artificial intelligence?based object detection techniques have been widely utilized throughout smart agriculture processes, including fruit maturity assessment and harvest time prediction. AI-based 3D object detection using LiDAR leverages these advantages to contribute to agricultural efficiency and the sustainable development of agriculture. In this study, we evaluate the performance of a 3D object detection model using point cloud data acquired by LiDAR in real agricultural environments.

https://doi.org/10.5370/KIEE.2026.75.3.629

박찬묵(Chan-Muk Park) ; 윤민호(Min-Ho Yoon) ; 임성훈(Sung-Hun Lim†)

With the increasing penetration of renewable energy, the detection of DC series arcs has become increasingly critical in DC power systems. This paper proposes an arc detection algorithm that uses frequency-band variations of current signals. Current data sampled at 100 [kHz] are processed using a sliding FFT (Fast Fourier Transform) and divided into seven bands spanning 0?50 [kHz]. The RMS current for each band is computed per frame, the absolute frame-to-frame difference is calculated, and the result is normalized using the Z-score method. An arc is declared if the metric exceeds a threshold within a detection window. DC series arc tests conducted under four load conditions and eight electrode opening speeds show pronounced changes in the 0.2?1 [kHz] and 1?5 [kHz] bands. The 1?5 [kHz] band, being less dependent on load yet sensitive to arcs, is selected for the algorithm. The proposed method detected arcs within 0.15 seconds across all tested conditions

https://doi.org/10.5370/KIEE.2026.75.3.637

김정기(Jeong-Gi Kim) ; 이진식(Jin-Sik Lee) ; 김재현(Jae-Hyun Kim) ; 김정환(Jeong-Hwan Kim) ; 최치우(Chi-Woo Choi) ; 전정채(Jeong-Chay Jeon) ; 김용혁(Yong-Hyeok Kim)

To overcome these limitations, this study applies three-dimensional X-ray computed tomography (3D X-ray CT) as a non-destructive technique for investigating electrical equipment failures. The CT system reconstructs multiple two-dimensional X-ray projections into a volumetric 3D model, allowing for the visualization and quantitative analysis of internal defects such as voids, cracks, carbonized zones, and melting marks without damaging the specimen. A series of case studies were conducted on various electrical components ? including power cables, straight joint splices, printed circuit boards (PCB), and vacuum circuit breakers (VCB). The results demonstrate that 3D X-ray CT enables clear identification of short-circuit-induced melting zones, insulation breakdown paths, and contact degradation areas that were previously unobservable by conventional methods. In particular, it allowed the determination of current flow direction, heat concentration patterns, and the progression of partial discharge-related damage in three dimensions. These findings highlight the effectiveness of 3D X-ray CT as a forensic and diagnostic tool in electrical safety engineering. Beyond post-accident analysis, the technique offers potential for predictive diagnostics and condition-based maintenance by providing reproducible, quantitative data on internal degradation. This study contributes to the advancement of scientific, standardized approaches to electrical failure investigation and establishes a foundation for integrating CT-based data analysis with artificial intelligence to enhance automation and objectivity in future safety diagnostics.

https://doi.org/10.5370/KIEE.2026.75.3.644

정인주(Inju Jung) ; 송진우(Jinwoo Song) ; 양주란(Jooran Yang) ; 김형표(Hyungpyo Kim)

This study aims to analyze the temporal variation of earth leakage current under moisture-exposed conditions and to evaluate the predictive performance of an LSTM-based time-series model. Unlike experiments intended to induce insulation degradation, the leakage environment was artificially simulated through water-contact and immersion to validate the model’s ability to detect pattern changes. Long-term leakage current data were collected from lighting-load conditions and applied to an LSTM network for future-current prediction. The model demonstrated high similarity to the measured data and accurately forecasted increases in leakage current up to the 7~9mA range. These results indicate that gradual leakage-current increases can occur under moisture exposure and that such patterns can be effectively learned by the LSTM model, although the findings do not constitute direct evidence of insulation degradation. The proposed model employs a lightweight architecture suitable for MCU-based embedded systems, enabling real-time processing in low-resource environments. This highlights its applicability to IoT-based monitoring platforms and its potential contribution to predictive maintenance frameworks for electrical installations.

https://doi.org/10.5370/KIEE.2026.75.3.651

백종길(Jong-Gil Baik) ; 구경완(Kyung-Wan Koo)

The growing demand for high-efficiency and low-carbon heating technologies has accelerated interest in alternative energy systems for aquaculture operations. This study investigates an induction heating boiler that integrates magnetic reluctance and electromagnetic induction to achieve enhanced thermal performance. Vegetable oil is employed as the thermal medium to improve heat stability, electrical insulation, and environmental sustainability. An 80-kW prototype was tested under real operating conditions in an inland seawater aquaculture facility. Experimental results showed that the system provided the same heating capacity of 160,000 kcal/h as a conventional electric boiler while consuming only 43% of the electrical power. The resulting relative thermal efficiency was measured at 232.5%, contributing to an estimated annual electricity cost reduction of approximately 55 million KRW and a CO₂ emission reduction of 365 tCO₂eq. The vegetable-oil medium maintained stable physical and electrical properties without significant degradation during continuous operation. These findings confirm that the proposed system offers high energy efficiency, operational safety, and low environmental impact. The results demonstrate strong potential for applying induction heating technology to industrial and aquaculture heating systems, particularly in support of national carbon-neutrality goals.

https://doi.org/10.5370/KIEE.2026.75.3.658

(Almas Saduakas) ; (Assel Mukasheva) ; (Alibek Bisembayev) ; (Dina Koishiyeva) ; 강정원(Jeong Won Kang)

Forecasting dynamics in financial markets remains a central yet unresolved challenge due to their inherent volatility, nonlinear dependence, and the presence of structural breaks and seasonality. The accurate modeling of stock price movements is not only of theoretical significance but also of considerable practical relevance for investment strategy design, portfolio optimization and systemic risk management. In this study, we investigated a comprehensive comparative analysis of forecasting techniques, encompassing both classical statistical models and modern machine learning approaches specifically adapted for time series prediction. Traditional econometric methods, as generalized autoregressive conditional heteroskedasticity (GARCH) and autoregressive integrated moving average (ARIMA) were systematically evaluated alongside conventional machine learning algorithms, including linear regression and support vector machines (SVM). Beyond these baselines, we assessed the predictive capacity of advanced neural network architectures, with particular emphasis on long short-term memory (LSTM) networks and convolutional neural networks (CNN), which are designed to capture long-range temporal dependencies and nonlinear feature interactions. Empirical experiments conducted on real stock market datasets allow for a rigorous performance assessment under diverse market regimes. The results demonstrated differentiated strengths across methods, where statistical models retain interpretability and robustness, while deep learning approaches yield superior accuracy in highly volatile environments. The study concludes with evidence-based recommendations concerning methodological suitability for varying forecasting horizons and financial application scenarios.

https://doi.org/10.5370/KIEE.2026.75.3.667

이동주(Dong-Ju Lee) ; 김종겸(Jong-Gyeum Kim)

This study analyzes the causes of power factor degradation in induction motors used in logistics systems and proposes effective improvement methods. Specifically, the impact of gear reduction ratios in conveyor systems on the motor’s power factor was simulated and evaluated. The results demonstrate that as the gear reduction ratio increases, the active power decreases significantly while the reactive power remains relatively constant, leading to a sharp decline in the overall power factor. Since a low power factor increases operating costs, reactive power compensation is essential to meet the standards required by power utilities (0.92 or higher). Furthermore, this study confirms that the required compensation varied by gear ratio and that the calculated values remains below the magnetization requirements, thereby eliminating the risk of self-excitation. These findings provide a crucial basis for diagnosing low power factor issues in logistics systems and developing safe, efficient compensation techniques.

https://doi.org/10.5370/KIEE.2026.75.3.673

어진명(Jin-Myeong Uh) ; 장진혁(Jin-Hyuk Jang) ; 손진근(Jin-Geun Shon)

This paper presents a method to improve the harmonic measurement and compensation performance of an active power filter (APF) by applying an individual Park Transform and designing an appropriate LPF. This method can improve accuracy, reliability and response speed more than applying a conventional LPF without continuously needing one cycle of data when measuring by using individual Park Transform. The simulation results show that when an appropriate LPF(butterworth filter) is designed using the proposed method, the voltage ripple is reduced compared to when a conventional LPF is used, and the response speed can be improved by at least 18.22 [%]. Accuracy is also validated by agreement with conventional FFT-based measurement results.The proposed method is expected to be applicable to harmonic-compensating devices such as APFs and hybrid filters, as well as harmonic measurement and power analysis equipment.

https://doi.org/10.5370/KIEE.2026.75.3.680

오성훈(Sung-Hoon Oh) ; 곽근영(Geun-Yeong Kwak) ; 이동우(Dong-Woo Lee)

This study quantitatively investigates the trade-off between power density and semiconductor loss as a function of switching frequency in a GaN-based four-switch non-inverting buck?boost converter for wearable-robot power applications and proposes a recommended switching frequency range. Well suited to mobile power systems with wide input-voltage variation, the non-inverting buck?boost converter supports non-inverted step-up and step-down operation, while the four-switch topology enables a single-inductor implementation that reduces passive components and footprint. When the input and output voltages are comparable, all four switches may commutate within one switching period, increasing loss; under high-frequency hard switching, switching loss can dominate and impose thermal and efficiency penalties. The converter is evaluated over an input-voltage range of 30 to 54.6 V with a 48 V/400 W output while sweeping switching frequency from 100 kHz to 1 MHz. Conduction and switching losses are evaluated separately to capture loss trends versus switching frequency and input voltage. To represent inductor burden consistently, an inductor-current-based energy proxy is defined over a steady-state time window. A Pareto plot using the energy proxy and total semiconductor loss identifies operating points that balance compactness and loss, providing a practical guideline for switching frequency selection.

https://doi.org/10.5370/KIEE.2026.75.3.690

손규연(Kyu-Yeon Son) ; 정형근(Houng-Kun Joung)

This paper proposes an intelligent automatic counting system leveraging a decentralized edge-server architecture to optimize micro-part enumeration in smart manufacturing. Unlike conventional weight-based or static 2D vision systems that suffer from occlusion and environmental sensitivity, the proposed system introduces a physical dispersion mechanism using high-frequency vibration and high-uniformity LED backlighting. This hardware-software synergy ensures robust object separation and maximizes visual contrast for a YOLOv8s-based detection engine. Experimental results demonstrate a counting accuracy of 92.4% and a processing throughput of 12.5 pieces per second, maintaining a stable 30.2 FPS transmission via a WebSocket-based data pipeline. The proposed approach overcomes the computational constraints of embedded devices while providing a scalable, high-speed solution for labor-intensive manufacturing processes.