https://doi.org/10.5370/KIEE.2025.74.8.1281

전승찬(Seungchan Jeon) ; 장문석(Munseok Chang) ; 배성우(Sungwoo Bae) ; 박성준(Seongjun Park)

This paper proposes a method for determining the optimal locations and capacity of multi-terminal DC (MTDC) systems in power grids with high renewable energy penetration. Using a master-slave structure for MTDC, DC voltage and power control are modeled. A multi-objective optimization problem is addressed using a voltage deviation profile index (VDPI), line power loss index (LPLI), and capacity penalty index (CPI). The optimization is solved through the artificial bee colony (ABC) algorithm. The method was validated in the IEEE-30 bus test grid. The results showed that an 18 MW MTDC system was optimal, with the DC voltage control converter at bus 30 and power control converters at buses 11 and 26. After MTDC integration, VDPI improved by 26.7%, and LPLI decreased by 19.0%. Additionally, the minimum voltage increased from 0.983 p.u. to 0.997 p.u. This highlights the impact that the locations and capacity of MTDC have on grid stability and efficiency.

https://doi.org/10.5370/KIEE.2025.74.8.1290

장현석(Hyeon-Seok Jang) ; 채정목(Jung-Mok Chae) ; 조윤성(Yoon-Sung Cho)

This study evaluates the impact of power facilities and standards on the stable operation of the Jeju power system, which comprises carbon-free power sources in alignment with the Carbon-Free Island (CFI) 2030 initiative. It analyzes the contributions of grid-forming energy storage systems (GFM-ESS), high-voltage direct current (HVDC) systems, synchronous condensers, primary frequency response, and fast frequency response to system stability. The study also investigates operational strategies for maintaining stability in scenarios without synchronous generators, incorporating grid stabilization resources. Furthermore, it identifies the required capacity of each stabilization resource to ensure the reliable operation of the Jeju power system.

https://doi.org/10.5370/KIEE.2025.74.8.1299

배규현(Gyu-Hyeon Bae) ; 윤아윤(Ah-Yun Yoon) ; 김성수(Sung-Soo Kim)

Renewable energy facilities are expanding globally, with photovoltaic (PV) systems rapidly increasing in South Korea. As PV systems grow, behind-the-meter (BTM) PV without real-time metering continues to expand, creating significant challenges for power system operation and forecasting. This study proposes a method to calculate BTM PV generation using market-traded PV performance and energy storage systems (ESS) charge/discharge power. By allocating ESS charge/discharge power regionally, the approach more realistically identifies BTM PV generation reflecting regional differences in generation characteristics. The analysis reflects regional generation characteristics more effectively compared to nationwide uniform approaches. This is expected to contribute to efficient and accurate power system operation and long-term planning in a power system environment where the proportion of BTM PV continues to increase.

https://doi.org/10.5370/KIEE.2025.74.8.1307

이광호(Kwang-Ho Lee)

This paper analyzes the effect of reliability improvement and congestion cost reduction through utilizing demand resources. In order to secure sufficient reserve margin, a new term for reserve margin is proposed into the social welfare index. The effect of load shifting and load leveling can be achieved by controling the reserve parameter according to the peak load or base load. Transmission congestion causes a difference in nodal prices on the congested line. The reduction of congestion cost can be achieved by giving differential incentives to the demand response aggregators according to the relative location based on the congested line.

https://doi.org/10.5370/KIEE.2025.74.8.1314

이기복(Ki-Bok Lee)

This paper presents the analysis on the characteristics of the IED based busbar protection systems based on the practical scheme designs. Any faults on the substation busbar mostly result in critical problems for system operation. Therefore, applying proper busbar protection scheme is very important. The types of busbar protection systems applied to the substations nowadays are normally high impedance scheme using OCR IED(Intelligent Electronic Device), dedicated high impedance busbar protection IED using current detection element scheme and IED based low impedance centralized & de-centralized scheme. The use of low impedance based busbar protection scheme becomes prevalent but the high impedance busbar protection schemes are still used since each type of busbar protection schemes has its own advantages and disadvantages. The proposed application guidelines of each scheme are presented using manufacturer data with detailed scheme analysis and comparison on design benefits.

https://doi.org/10.5370/KIEE.2025.74.8.1324

박영호(Young-Ho Park) ; 김수현(Su-Hyeon Kim) ; 임성훈(Sung-Hun Lim) ; 이종필(Jong-Pil Lee)

In a power distribution system supplying 6.6 kV to multiple feeders connected to various loads, a fault on one feeder can result in voltage drops across other connected feeders. Voltage sags can severely affect on facilities with rotating machinery by causing motors to draw excessive current to maintain output power and continuous or repeated voltage sag lead to overheating, insulation damage and motor failure. This paper studied protection coordination between under voltage relay (UVR) in a sound feeder and overcurrent relay (OCR) in a faulted feeder considering application of a superconducting fault current limiter (SFCL). Proper current limiting reactor/resistor (CLR)’s value comprising the SFCL to ensure the coordination between OCR and UVR was investigated to prevent relay malfunction. Through the comparative analysis due to the impedance magnitude of resistive and inductive CLRs, the proper CLR value of the SFCL was confirmed to be required to protect sensitive rotating machinery load.

https://doi.org/10.5370/KIEE.2025.74.8.1332

이경민(Kyung-Min Lee) ; 박철원(Chul-Won Park)

In this paper, in order to improve the reliability of IED operation by relay signals due to the saturation of CT magnetic material, a transient analysis considering nonlinear CT saturation of the three phase three winding transformer system is performed. First, the Jiles-Atherton model equation for nonlinear CT magnets is formalized. After selecting a domestic actual three phase three winding transformer system, the related system parameters are collected. We model a CT based three winding transformer system using PSCAD, an EMT tool, and verify its validity in steady state. Finally, the hysteresis curves and transient currents of the magnetizing inrush and fault state are analyzed and evaluated by performing simulations under various conditions.

https://doi.org/10.5370/KIEE.2025.74.8.1341

정태욱(Tae-Uk Jung)

In industrial sites for power facility construction and management, if there is a safety device that indicates whether the power is on, it can prevent electric shock accidents and make electrical work and maintenance inspections safer. This paper applies a technology method that collects electromagnetic energy drifting around power lines through a receiver by applying magnetic energy harvesting technology. We develop an energy indicator that displays the energy status of power lines using a high-brightness LED with its own power source without a separate external power source such as a battery by utilizing the collected electric energy. In this paper, we examine the modeling and analysis methods of a harvesting system using magnetic energy, and derive the design results of an energy collector that satisfies the required specifications.

https://doi.org/10.5370/KIEE.2025.74.8.1348

김동현(Dong-Hyeon Kim) ; 안태영(Tae-yong Ahn) ; 김효준(Hyo-jun Kim) ; 이호준(Ho-Joon Lee)

Eddy current losses generated in permanent magnet motors during high-speed operation cause internal heating of the magnets, leading to a temperature rise that can result in irreversible demagnetization. This phenomenon directly affects the output stability and reliability of the motor. In particular, high-performance materials such as NdFeB-based rare-earth magnets exhibit high temperature sensitivity, requiring careful consideration during the design process. A common method for suppressing eddy current losses is to segment the magnets, thereby interrupting the path of the induced currents. However, such physical segmentation increases the complexity and cost of precision machining and assembly, and presents significant challenges when applied to axial flux motors. In this study, an integrated magnet structure is proposed that effectively controls eddy current paths without requiring segmentation. The proposed design is evaluated using three-dimensional finite element analysis to compare torque characteristics, induced current distribution within the magnets, and the resulting eddy current losses with those of a conventional design.

https://doi.org/10.5370/KIEE.2025.74.8.1354

송인혁(In-Hyuk Song) ; 한성희(Seong-Hee Han) ; 노윤섭(Youn-Sub Noh) ; 임종원(Jong-Won Lim) ; 권지훈(Ji-Hun Kwon) ; 전병철(Buoung-Chul Jun) ; 김동욱(Dong-Wook Kimark)

In this paper, two types of X-band GaN low-noise amplifier(LNA) monolithic microwave integrated circuits(MMICs) are designed, fabricated, and evaluated using a 0.2 μm GaN HEMT process at Wavice foundry. The low-noise amplifier MMICs use GaN HEMTs with source degeneration microstrip lines, and insert thin-film resistors in shunt (shunt RC) or series configurations into the biasing networks for amplifier stabilization. Input matching of a two-stage amplifier MMIC is implemented to obtain a lower noise figure, and input matching of a three-stage amplifier MMIC is achieved for higher gain and better input return loss. The fabricated two-stage MMIC exhibits a linear gain of 16.2~17.3 dB, a noise figure of 1.31~1.52 dB, and an output P1dB of 26.5 dBm, while the three-stage MMIC achieves a linear gain of 22.5~24.7 dB, a noise figure of 1.67~1.77 dB, and an output P1dB of 25.5 dBm. Both LNA MMICs have a saturated output power of 32.6 dBm. Furthermore, the power measurements show that the three-stage MMIC is endurable up to the input power of 10 W. The MMIC chip dimensions are 1.6×1.5 mm2 for the two-stage amplifier and 1.85×1.5 mm2 for the three-stage amplifier.

https://doi.org/10.5370/KIEE.2025.74.8.1363

임창석(Changseok Lim) ; 주도윤(Doyoon Ju) ; 이영삼(Young Sam Lee)

This paper proposes the implementation of equilibrium-to-equilibrium transition control for a triple inverted pendulum system using Sim-to-Real reinforcement learning. Recently, multi-link inverted pendulum systems have introduced the new control challenge, extending beyond conventional swing-up and balancing controls toward equilibrium-to-equilibrium transition control. Transition control, which involves continuous transitions between multiple unstable equilibrium points, is particularly sensitive to disturbances. To address this, we apply the Sim-to-Real reinforcement learning technique, transferring control policies learned in simulation to the physical system. Furthermore a triple inverted pendulum system with high model consistency was designed and constructed to minimize the reality gap between simulation and physical environments. The proposed controller successfully achieved all 56 possible transitions among the eight defined equilibrium points. The results demonstrate that transition control based on Sim-to-Real reinforcement learning effectively resolves complex nonlinear control problems.

https://doi.org/10.5370/KIEE.2025.74.8.1373

윤주원(Joowon Yun) ; 김은빈(Eunbin Kim) ; 이언석(Onseok Lee)

Proprioception is the fundamental sense that enables the central nervous system to process body positions, postures, and movements in functional tasks. Aging and neurological disorders often lead to proprioceptive impairments that affect balance and coordination. This study developed a contactless system using Kinect and Unity 3D to evaluate proprioception by quantifying joint angles under visual and positional restrictions. Participants performed specific postures with proprioceptive limitations induced by 2 kg weights and closed-eye conditions. Paired t tests revealed significant proprioceptive variations under these constraints. Random forest achieved a classification accuracy of 96%, effectively distinguishing sensory changes across conditions. These results highlight the potential of this system for rehabilitation and training programs targeting proprioceptive function.

https://doi.org/10.5370/KIEE.2025.74.8.1381

박이형(Yi-Hyeong Park) ; 이현용(HyunYong Lee) ; 강창묵(Chang Mook Kang)

Power Factor Correction (PFC) circuits play a vital role in improving power quality and ensuring the stability of power systems. However, collecting real-world fault data for these circuits is costly and time-consuming, making it difficult to train reliable diagnostic models. To address this issue, this study proposes a data augmentation method using a Transformer-based Generative Adversarial Network(GAN) integrated with Positional Encoding. The proposed approach captures the temporal dependencies and nonlinear characteristics of PFC fault signals more effectively than traditional techniques. Experimental evaluations using t-SNE, Maximum Mean Discrepancy(MMD), and multiple classification models confirm the advancement of the proposed method in generating realistic and diverse fault data. This research contributes to enhancing the robustness and accuracy of fault diagnosis models and offers scalability to other power electronic systems.

https://doi.org/10.5370/KIEE.2025.74.8.1389

김규식(Kiu-Sik Kim) ; 심덕선(Duk-Sun Shim)

An optimization algorithm is essential for minimizing loss (or objective) functions in machine learning and deep learning. This article compares empirically the performance of 8 optimization algorithms which are used for training in machine learning and deep learning. The optimization algorithms considered in this article consist of 6 well-known algorithms such as SGD, SGD momentum, Nesterov momentum, AdaGrad, RMSprop, and Adam algorithms, and 2 recent algorithms such as KO and mSGD algorithms. Three kind of data is used for the performance comparison of optimization algorithms. First is the use of two functions, which have many local minima. Second is the use of MNIST data set which consists of handwritten digits from 0 to 9. The last one is the use of CIFAR-10 which consists of 10 kinds of images such as airplane, car, cat, dog, and so on. With the three kind of data set, seven cases are considered including FNN, CNN, and AE to compare the performance of 8 optimization algorithms.

https://doi.org/10.5370/KIEE.2025.74.8.1399

김길중(Gil-Jung Kim) ; 이순명(Soon-Myung Lee)

The adoption of Electric Vehicles (EVs) is increasing globally. However, the EV market has recently entered a chasm phase, slowing growth. Surveys and Research indicate that challenges not only related to EVs themselves but also to DC electric vehicle supply equipment (DC EVSE) act as major barriers. This study focuses on compatibility and safety issues in DC EVSE and explores solutions based on international standards. Using compatibility and safety as classification criteria, this study identifies nine major changes introduced in IEC 61851-23:2023. These changes have not yet been reflected in domestic standards. However, incorporating them into future revisions could improve compatibility and safety, thereby enhancing the reliability of DC fast charging infrastructure. Ultimately, such standardization is expected to accelerate EV adoption and facilitate the expansion of charging infrastructure.

https://doi.org/10.5370/KIEE.2025.74.8.1412

김종겸(Jong-Gyeum Kim)

Induction motors, as typical inductive loads, have the characteristic of lagging power factor because of the reactive power required to generate the rotating magnetic fields. Even when an induction motor is driven with a load close to the rated output, the power factor is mostly around 0.8 due to the influence of the inductive component. Recently, the importance of power factor has increased as power companies have increased the power factor from the existing 0.9 to 0.92. Common power factor compensation approaches include individual correction for each motor, group compensation for multiple motors, correction at the transformer’s secondary side, or a combination of these methods. This study simulated and analyzed the changes when power factor compensation capacitors were installed on the motor, transformer, and both sides under conditions where the load varied in the field and the size of the active power was different. As a result of the analysis, it was confirmed that it is desirable to install power factor correction devices of an appropriate size on both sides, and in particular, the size of the reactive power installed in the transformer needs to be calculated according to no-load operation.

https://doi.org/10.5370/KIEE.2025.74.8.1418

이해원(Hae-Won Lee) ; 최장영(Jang-Young Choi)

The form of the railway track includes open route, bridge, tunnel, and even tracks close to the coast. Therefore, insulators for catenary are exposed to more dust, salt, and moisture than general electrical equipment installed on roads. Considering this environment, insulators for AC 25kV catenary lines are designed and applied with an insulation class of 46kV and an specifie creepage distance(SCD) of Very Heavy pollution level exceeding 25mm/kV. However, there have been 185 cases of catenary insulator flashover in the past 13 years. In this paper, we analyzed cases of flashover caused by dust and moisture in tunnels, flashover caused by salt damage, and flashover of polymer insulators in high-speed railway open route, and examined the effect of railway environment on the insulation performance of catenary insulators.

https://doi.org/10.5370/KIEE.2025.74.8.1426

좌희정(Heejung Jwa) ; 정문희(Munhee Jeong) ; 조정원(Jungwon Cho)

In this study, we have assessed the zero-shot classification performance of Jeju Island plant images using five multimodal vision?language models: CLIP, SigLIP, SigLIP Multilingual, SigLIP SO400M, and SigLIP2. Evaluation data comprised image?text pairs of plant species collected from four ecologically distinct regions (Deonggae Coast, Min-oleum, Jabaebong, and Jeju City). All models were evaluated under an identical zero-shot classification protocol to ensure a fair comparison. Among them, SigLIP SO400M achieved the highest accuracy on the Deonggae Coast subset, attaining a macro accuracy of 0.7460 and a micro accuracy of 0.7612, thereby outperforming the other models. The prompt language format exerted a significant influence on performance: English-only prompts consistently surpassed Korean-only prompts across all models. Confusion matrix analysis revealed region-specific class-level misclassification patterns, identifying species prone to frequent confusion. Collectively, these results demonstrate the robust zero-shot classification capabilities of contemporary vision?language models for fine-grained plant species identification and underscore the importance of selecting both an appropriate model and prompt format for a given task. The code used for these experiments is publicly available at github.com/flyaround365/JejuPlantsClassification.