https://doi.org/10.5370/KIEE.2025.74.4.539

박채운(Chaewoon Park) ; 전영신(Youngshin Jeon) ; 노재형(Jae Hyung Roh)

In response to the climate crisis, the expansion of renewable energy has accelerated. However, due to the intermittency and uncertainty of these resources, it has become difficult to maintain real-time supply-demand balance solely based on the generation plans from the day-ahead market. As a result, real-time market implementation is being introduced. Additionally, the grid's stability is declining under these conditions, and the importance of flexible resources is increasing. In this context, district energy providers, who operate combined heat and power(CHP) plants capable of offering flexibility, play a crucial role. However, district energy providers face challenges in formulating their strategies in the day-ahead and real-time markets due to their obligation to supply heat, making it more complex compared to other GENCOs. Therefore, this paper focuses on the bidding strategies of district energy providers in this evolving market environment.

https://doi.org/10.5370/KIEE.2025.74.4.547

김윤호(Yun-Ho Kim) ; 이민행(Min-Haeng Lee) ; 김경화(Kyung-Hwa Kim) ; 장형안(Hyeong-An Jang) ; 최성문(Sung-Moon Choi) ; 노대석(Dae-Seok Rho)

The research on the LVDC distribution system is being conducted as the renewable energy resources and DC load has been rapidly increased. The LVDC distribution system has advantage of utilization of DC energy resources and reduction of power loss, but the research on stable operation and protection coordination methods are being required. In particular, when the fault is occurred in the LVDC distribution system, the interruption may be extended due to shutdown of main converter. Therefore, This paper deals with the formulation of slope characteristics of fault current considering line constants, customer load, internal impedance of section converter and presents the control method to detect and evaluate the fault in accurate manner in intelligent protection coordination device(IPCD). And also, this paper performs the modeling of the IPCD using PSCAD/EMTDC and then implements the IPCD which is composed of H/W power section and S/W control section. From the simulation and test results based on the proposed formulation of fault current, the control method of IPCD is expected to be effective tool because the slope characteristics of PSCAD/EMTDC modeling is nearly equivalent with one of H/W device. And also, it is found that the proposed IPCD with the slope characteristic of fault current can evaluate the fault conditions and limit the fault current in fast and accurate manner compared to the existing method with the magnitude of fault current only. Therefore, it is confirmed that the proposed IPCD can minimize the outage area in LVDC distribution system by preventing the shutdown of main converter.

https://doi.org/10.5370/KIEE.2025.74.4.561

김종겸(Jong-Gyeum Kim)

Induction motors are widely used as a power source for rotating loads due to their simplicity of structure, robustness, long lifespan, and ease of maintenance. However, these motors can cause voltage drops in loads connected to the same line when a significant voltage drop occurs during the starting process, as high current is generated. To mitigate this issue, various starting methods are employed based on the motor’s capacity to reduce the starting current. In drainage facilities that are operated temporarily during the rainy season, low-head pumping facilities are typically used. For such drainage pumping stations, which require the instantaneous discharge of large volumes of water, high-capacity multi-pole induction motors are often employed. When the motor capacity is large, reactor starting is frequently chosen over direct starting. Additionally, it is essential to consider the duration of the voltage drop when selecting the starting method. In drainage facilities, the facility capacity may be large due to temporary operations, but the operating system is relatively simple. This study compared and analyzed changes in voltage and the time required to reach the rated speed when pumping equipment, considering the moment of inertia, was started directly using a multi-pole induction motor and when started with a reactor. The analysis confirmed a difference in response time, and the voltage drop was found to be within the allowable range.

https://doi.org/10.5370/KIEE.2025.74.4.568

변범석(Beom-Seok Byeon) ; 박의종(Eui-Jong Park) ; 김용재(Yong-Jae Kim)

In the magnetically geared machine with an inner stator(GIS) type, a magnetic geared machine, it has been observed that a significant discrepancy arises between the motor's maximum output torque and the gear's maximum input torque when the gear ratio of the magnetic gear is low. This issue persists even when the final output through the magnetic gear remains the same, due to the extensive use of permanent magnets and iron in the magnetic gear. Therefore, this paper aims to explore a method for selecting the torque density of GIS-type magnetic geared machines(MGM) by intentionally reducing the torque density of the magnetic gear.

https://doi.org/10.5370/KIEE.2025.74.4.575

박지훈(Ji-Hoon Park) ; 이민영(Min-Young Lee) ; 강동우(Dong-Woo Kang) ; 김태형(Tae-Heoung Kim)

The electric vehicle market and the motor market have recently been focusing on lightweight designs and high specific power output. However, there are limited ways to enhance the specific power output of Radial Flux Motors (RFM), commonly used motors. To overcome the limitations of improving the specific power output of motors like RFMs, this study explored a method to increase specific power output by reducing the weight of the windings using Copper Clad Aluminum (CCA) windings. CCA windings are aluminum windings coated with copper, which offer advantages such as lower weight and cost compared to copper windings, and higher electrical conductivity and strength compared to aluminum windings. This study utilized Ansys Electronics Desktop 2024 R1, an electromagnetic analysis software using the finite element method, to compare the copper losses, total losses, efficiency, and specific power output of two types of CCA windings with pure copper windings, thereby validating the proposed method.

https://doi.org/10.5370/KIEE.2025.74.4.581

이상욱(Sang-Uk Lee) ; 문주형(Ju-Hyeong Moon) ; 김태형(Tae-Hyeong Kim) ; 강동우(Dong-Woo Kang)

Radial flux motors (RFM) have achieved significant technological advancements to meet the high-power demands in various industries. However, due to their structural limitations, research on axial flux motors (AFM) has been actively progressing. Among them, yokeless and segmented armature axial flux (YASA-AFM) motor can generate high torque proportional to the motor diameter by utilizing its structural characteristics, thus offering enhanced output performance. Despite the ongoing research on the design of dual-rotor structures in YASA-AFM, there is a lack of clear criteria for efficient rotor diameter design to achieve high power output relative to volume. Furthermore, the air gap flux generated by the two rotors can cause flux leakage due to magnetic saturation in the single stator, potentially reducing the motor's output. To address these issues, this paper aims to study the rotor structure characteristics for minimizing flux leakage in AFM and provide criteria for efficient rotor diameter design. The analysis and validation of this study will be conducted using the finite element analysis software ANSYS Electronics Desktop 2024 R1. Through this research, we aim to improve the design of AFM and expand its applicability to a broader range of industrial fields.

https://doi.org/10.5370/KIEE.2025.74.4.586

최동훈(Dong-Hoon Choi) ; 이동주(Dong-Ju Lee)

A multidisciplinary design optimization (MDO) was performed for a rectifier, which is one of key components of uninterruptible power supplies (UPS), using AI-Aided Design Optimization (AADO) technology. First, the design variables and performance indices of the target rectifier were clearly defined to formulate a MDO problem. Then, a multidisciplinary analysis procedure was established using commercial simulation software. Through AADO technology, optimal circuit design variables and heatsink shape parameters were successfully determined, achieving a 36.1% reduction in heatsink volume while satisfying various constraints on circuit performance and temperature.

https://doi.org/10.5370/KIEE.2025.74.4.594

박성현(Sunghyeon Park) ; 김인동(In-Dong Kim)

Recently, as the demand for DC systems has rapidly increased, research on battery chargers has been conducted extensively. Among DC/DC converters, the DAB converter has the advantage of enabling ZVS(Zero Voltage Switching) operation and securing bidirectional power transmission and insulation by using a transformer in a left-right symmetrical structure. Since bidirectional operation is possible, it is widely used in DC grid systems and battery chargers. This paper proposes a switching method that integrates three switching methods, TPS, EPS, and SPS, for efficient converter operation in a wide load range. In addition, a controller that smoothly operates CC-CV (Constant Current Constant Voltage) mode conversion and has self-protection operation is ? proposed, and the operating characteristics are verified through simulation and experiments using the fabricated DAB converter.

https://doi.org/10.5370/KIEE.2025.74.4.605

최원정(Won-Jeong Choi) ; 김상훈(Sang-Hoon Kim)

Voice Coil Motor(VCM) has many advantages such as no mechanical backlash, fast response, low power driving and small size, etc. This motor is widely used in fields that require high precision position control, so high performance position control is essentially required. This paper suggested a new method of designing PID feedback position controller gains. But, there is a limit to improving the position control precision with only feedback controller, so feedforward control is essential. Feedforward terms required in a VCM include inertial force, frictional force and spring force. Among the three terms, frictional force in particular is a nonlinear element that hinders the precise position control of a VCM. Therefore, this paper proposed a new friction model that considers nonlinear characteristics based on friction data obtained through experiments. By applying the designed PID position controller gains and a proposed model, in experiments, position precision improvement will be demonstrated through position error.

https://doi.org/10.5370/KIEE.2025.74.4.614

박준휘(JunHwi Park) ; 마재성(Jaesung Ma) ; 김동윤(Dong-Youn Kim) ; 강정화(Junghwa Kang) ; 김영민(Youngmin Kim) ; 류현정(Hyun-jung Ryu) ; 윤동원(Dong Won Yoon) ; 김장목(Jang-Mok Kim)

This paper presents a method to enhance the reliability of a dual 3-phase IPMSM (Interior Permanent Magnet Synchronous Motor) system with a parallel stator-winding structure by introducing redundancy in the motor, inverter, and rotor position sensors. The system includes two rotors on a single shaft, managed by one speed controller and two current controllers, allowing one motor to compensate if the other fails. Inverter redundancy ensures synchronized control and continued operation if one inverter fails. A sensorless control algorithm is implemented for emergency operation during position sensor failure, using resolver-based hardware and software-based sensorless fault detection. The IF(Current Frequency) - sensorless control method provides stable emergency ? operation both at rest and in motion. The paper also introduces an improved speed controller for smooth transitions, with experimental results validating the algorithm's effectiveness.

https://doi.org/10.5370/KIEE.2025.74.4.624

성정원(Jung-Won Sung) ; 이은규(Eun-Gyu Lee) ; 곽상신(Sangshin Kwak)

As capacitors are under age, their performance degrade, which can lead to reduced reliability of the power converter. In this study, we propose a method for effectively analyzing capacitor’s condition by combining a data preprocessing technique using Lissajous curves with an AI model. The experimental results confirmed that the proposed method accurately predict the aging state of the capacitor by using the proposed Lissajous curves based preprocessing approach, demonstrating performance even under duty conditions that had never been trained on.

https://doi.org/10.5370/KIEE.2025.74.4.629

김재원(Jae-Won Kim) ; 박수성(Su-Seong Park) ; 박해찬(Hae-Chan Park) ; 김래영(Rae-Young Kim)

As the demand for electric vehicles continues to rise, the development of a high-power ultra-fast charging system with a wide output voltage range is essential to reducing battery charging time and extending driving range. The single-phase LLC converter, which achieves zero voltage switching over the entire operating range, is widely used as a DC-DC converter in electric vehicle charging systems. However, in high-power applications, increased input and output current ripple leads to higher system losses and larger volume. To address these challenges, a three-phase interleaved LLC converter is proposed to reduce input and output current ripple while enhancing system efficiency and power density. This paper presents the resonant network design of a 30 kW three-phase interleaved LLC converter with a wide output voltage range, intended for ultra-fast EV charging applications. The performance of the proposed system is verified through a prototype, achieving a maximum efficiency of 98.24% and a power density of 3,086 W/L.

https://doi.org/10.5370/KIEE.2025.74.4.635

박명근(Myeonggeun Park) ; 이학준(Hakjun Lee) ; 권우경(Wookyong Kwona)

This study proposes an integrated framework to enhance the autonomy and user interaction of serving robots. By combining spatial information processing, speech recognition, and large language models (LLMs), the system improves efficiency and usability in complex environments. It enables precise task execution and enhances user interaction through intuitive speech commands and natural language understanding. The framework also boosts operational efficiency, allowing robots to navigate autonomously and respond to commands in real-time. This approach advances both robot autonomy and user experience, offering a new paradigm in service robotics.

https://doi.org/10.5370/KIEE.2025.74.4.644

이종윤(Jong-Youn Lee) ; 서정윤(Jeong-Yun Seo) ; 박성준(Sung-Jun Park) ; 이하림(Harim Lee) ; 이용환(Yong-Hwan Lee)

In this paper, we introduce the design of an artificial intelligence (AI) hardware accelerator for a convolutional neural network (CNN) for classification tasks. The architecture of the AI hardware for a CNN model consists of two modules: a striding-padding (SP) module and a convolution layer module. For the classification task, fully-connected layers should follow the CNN model, and hence we also briefly explain the architecture of the fully-connected layer. In the architecture of a CNN model, the SP module manages how to access memories including feature maps and weights by considering the striding and padding operation of a convolution layer, and the convolution layer conducts convolution operation with feature maps and weights that are read from the memories. To verify the AI hardware design through RTL simulation, CNN models are trained in software using a MNIST dataset and a Fashion MNIST dataset. Then, post-training quantization and quantization-aware training, which are representative quantization techniques, are utilized to extract hardware-appropriate weights from the trained CNN models. The quantized weights are loaded to memories, and the validation data from the dataset used to train CNN models are fed to the designed CNN hardware. Finally, the inference results of the hardware are compared with that of the software-trained CNN models, which confirms the correctness of the hardware design.

https://doi.org/10.5370/KIEE.2025.74.4.655

서의석(Ui-Suk Suh) ; 나원상(Won-Sang Ra) ; 서태일(Taeil Suhrk)

This paper presents a method that analyzes long-term position errors in the SDINS(strap-down inertial navigation system) for underwater vehicle applications. Under the assumption that the initial platform misalignment and the inertial sensor biases are the primary sources of the position errors in the SDINS, an analytic solution is derived based on the error propagation model in which these factors serve as the state variables. By using the vertical channel damping loop within the SDINS, the analytic solution is obtained by focusing on the position error in the horizontal plane. The proposed method enables a mathematical analysis of the SDINS position errors with respect to each error state, observing the oscillation terms such as the Schuler/Foucault frequency and the Earth’s rotation. Thus, the proposed solution is expected to be used as a theoretical basis for the development of the underwater vehicles’ navigation system. The effectiveness of the proposed solution is verified by the results of SDINS position error through computer simulations.

https://doi.org/10.5370/KIEE.2025.74.4.666

정선교(Seon-Gyo Jeong) ; 강지훈(Jihoon Kang) ; 한광희(Kwang-Hee Han) ; 조관필(Kwanphil Cho) ; 나원상(Won-Sang Ra)

This paper proposes a precise skip trajectory prediction algorithm of a maneuvering reentry vehicle(MaRV). It is well known that most MaRVs tend to fly the skip trajectory while maintaining a certain lift-to-drag ratio to ensure the aerodynamic efficiency. This fact motivates us to derive a closed-form solution of the skip trajectory which expressed in terms of the lift-to-drag estimate. It shows that although the magnitude of the MaRV’s flight path angle are the same when it enters and exits the atmosphere, their signs are opposite. Exploiting this analysis result and the lift-to-drag estimates, we can efficiently make a long term prediction of the skip trajectory. Unlike existing methods, the proposed method does not require the establishment of a database of aerodynamic properties of MaRVs and reduces the computational burden of numerical integration, which is very suitable for practical applications. Through simulations on typical MaRV trajectory scenarios, it is verified that the proposed method provides superior skip trajectory prediction performance compared to existing schemes.

https://doi.org/10.5370/KIEE.2025.74.4.677

김한규(Hann-Gyoo Kim) ; 이승환(Seung-Hwan Lee)

This paper proposes a novel map merging technique designed to improve accuracy and reliability in environments with multiple dynamic objects by using temporal maps and probabilistic reliability maps. Traditional map merging methods often experience reduced accuracy in such settings due to the significant spatial and temporal variability introduced by dynamic objects. The proposed method addresses this by recording object update times in a temporal map and evaluating object reliability over time with a probabilistic reliability map, enabling effective handling of uncertainties caused by dynamic objects. In particular, the probabilistic reliability map estimates object reliability over time, allowing for the selective exclusion of less reliable data and ensuring a more precise merged map. Simulation results demonstrate that the proposed method outperforms existing approaches across various environments and scenarios, both visually and quantitatively. Future research will focus on extending this technique to 3D environments and validating its performance in real-world settings to enhance its practical applicability.

https://doi.org/10.5370/KIEE.2025.74.4.683

서장호(JangHo Seo) ; 이준우(Joonwoo Leerk)

In recent years, robotics technology has made significant advancements, particularly in multi-robot systems where efficient task allocation plays a crucial role in maximizing productivity and minimizing operational time. Previous research has explored various approaches to solving the Multi-Robot Task Allocation problem, but many have faced challenges in task distribution efficiency. To address this issue, we propose a Seeding Genetic Algorithm based on the A* algorithm and DBSCAN clustering. The A* algorithm performs path optimization in a grid environment with obstacles, while DBSCAN clusters tasks to enhance efficient task allocation. By seeding GA with these optimized solutions, the algorithm achieves faster convergence and higher solution quality. Simulations conducted on two maps with different robot configurations show that the A*-DBSCAN Seeding GA outperforms traditional GA and Greedy methods. The proposed method reduced the makespan, and its statistical significance was verified through ANOVA tests. This research contributes to improving multi-robot collaboration in industrial applications, offering an effective solution to the MRTA problem, reducing task completion time, and enhancing system efficiency.

https://doi.org/10.5370/KIEE.2025.74.4.691

남의석(Eui-Seok Nahmrk)

In this paper, we proposed a data quality and diversity guarantee plan and a method to verify the expert operation experience using RNN and a data quality and diversity guarantee plan to 1) solve the two problems of the sewage treatment expert driving experience and 2) the limitation of situation recognition and judgment. As a solution to the lack of data quality and diversity, not all data are used, but a method of filtering and using only the part that is the basis for determining the expert's experience was proposed. In addition, RNN was used to solve the problem of the limitation of situation recognition and judgment. The reason why RNN is used among various neural networks is that RNN is a neural network architecture for deep learning that predicts time series or sequential data, and sewage treatment data is also time series or sequential data. According to the proposed technique, a RNN model with a reliability of more than 8% was implemented, and the effectiveness of the proposed technique was verified by applying two experiences and one incorrect experience of actual operation experts of the target sewage treatment plant. As a result of the verification, both the actual driving expert's experiences 1 and 2 were verified as valid, and for incorrect experiences, there were 3,442 data that matched the expert's experience and 4,444 data that did not match, so the proposed method can verify the expert's experience.

https://doi.org/10.5370/KIEE.2025.74.4.700

최승규(Seung-Kyu Choi) ; 이경주(Kyoung-Joo Lee) ; 곽지혜(Ji-Hye Kwak) ; 김유민(Yu-Min Kim) ; 이재복(Jae-Bok Lee) ; 강성만(Sung-Man Kang)

In this paper, a MOV(Metal Oxide Varistor), which is a component of a MOSA(Metal Oxide Surge Arrester) for DC circuit breakers required in MVDC(Medium Voltage Direct Current) distribution networks, is proposed. The developed MOV has a reference voltage of 7.2 kV @ 1 mA, a residual voltage of 17 kV @ 10 kA, and an absorbing energy of 39 kJ. Therefore, the proposed MOV has suitable characteristics for configuring a MOSA for DC circuit breakers in 42 kV MVDC distribution networks which requires an absorption energy of 2 MJ. In this paper, the design process and performance test results of the proposed MOV are presented.

https://doi.org/10.5370/KIEE.2025.74.4.705

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

There are a total of 17 types of 25 kV AT-type catenary insulators, which can be categorized into suspension, stem, and support insulators. Their materials include glass, ceramic, and polymer. In the first phase of the Gyeongbu High-Speed Railway, insulators of a 46 kV insulation class made of glass were installed for the 27.5 kV rated system. In contrast, in the second phase of the Gyeongbu High-Speed Railway and in the Honam High-Speed Railway catenary, newly developed polymer insulators are adopted. In particular, the T-mx polymer insulator was designed to maintain the overall dimensions of the existing glass insulator while enhancing electrical performance for the movable bracket. However, railway operations in the field have requested further improvement in the polymer insulator’s insulation performance to mitigate frequent flashover accidents. Therefore, this paper examines the type indication of AC 27.5 kV catenary insulators and reviews the suitability of the insulation distance, including creepage distance, for a 46 kV polymer stem insulator.

https://doi.org/10.5370/KIEE.2025.74.4.712

윤치명(Chi-Myeong Yun) ; 민명환(Myung-Hwan Min)

In this paper, we analyzed the change in energy efficiency according to the output voltage of the substation before applying a device capable of active control such as electronic OLTC to the power traction system. The energy efficiency improvement for each situation was simulated by assuming two situations when the voltage was raised and when the voltage was lowered. Two railway operators compared the results of field measurement tests with the results using PSCAD/EMTDC, a power analysis program, to prove the improvement of energy efficiency according to voltage changes.

https://doi.org/10.5370/KIEE.2025.74.4.717

김영승(Young-seoung Kim) ; 나건우(Kun-woo Na) ; 서유정(U-jeong Seo) ; 김우용(Wooyong Kim)

Lithium-ion batteries have established themselves as a critical energy source for electric vehicles and energy storage systems. As their usage continues to grow, the issues of battery currents due to current sensor failures can lead to overcharging, over-discharging, and potentially hazardous incidents such as battery fires. To address this issue, this study proposes a current sensor fault detection algorithm based on the analysis of the data covariance matrix. The proposed algorithm is designed as an add-on type solution, enabling calculations using only voltage data extracted from the Battery Management System (BMS). It detects current sensor faults in real-time, thereby mitigating potential risks. Furthermore, the study demonstrates the algorithm’s applicability by cross-validating its performance using data from both new and aged battery models. Through this approach, the algorithm addresses current sensor fault issues across various battery states, contributing to the prevention of critical incidents such as battery fires.

https://doi.org/10.5370/KIEE.2025.74.4.727

이재건(Jae-gun Lee) ; 김진식(Jin-sik Kim) ; 이현재(Hyun-jae Lee) ; 손진근(Jin-geun Shon)

This study addresses power quality degradation issues caused by short-circuit faults in 3/2-phase switching inverters by estimating and applying an optimized voltage control range. The ripple effects, including voltage distortion, power factor reduction, and efficiency loss due to the reduction in maximum phase voltage during a short-circuit event, were analyzed using PSIM simulations. Based on these analyses, the application of a voltage control range for the DC-link capacitor effectively reduced phase voltage waveform distortion, improved overmodulation issues, and maintained power factor and apparent power. Consequently, this study provides a solution to enhance the stability and reliability of renewable energy systems while maintaining power quality and maximizing energy efficiency.

https://doi.org/10.5370/KIEE.2025.74.4.733

임종호(Jong-ho Lim) ; 송성근(Sung-geun Song) ; 이현재(Hyun-jae Lee) ; 손진근(Jin-geun Shon)

This paper proposes an Maximum Power Point Tracking(MPPT) control method utilizing a binary search algorithm to enhance the power generation efficiency of photovoltaic(PV) systems. In PV systems, the MPPT control is a technique designed to maintain power generation at the maximum power point (MPP) under all conditions. The MPPT control must quickly track the MPP in real-time in response to fluctuations in power generation caused by external factors. To achieve rapid and accurate identification of the maximum power point, this study applies a binary search algorithm, which efficiently finds a target value within a sorted array, to the MPPT control. The performance of the proposed MPPT control method was validated through simulations conducted using PSIM. The results demonstrate that the proposed approach effectively increases the power generation efficiency of PV systems.

https://doi.org/10.5370/KIEE.2025.74.4.739

노민수(Min-Soo Noh) ; 김동현(Dong-Hyeon Kim) ; Youguang Guo(Youguang Guo) ; 이호준(Ho-Joon Lee)

Magnetic bearing motors are in high demand due to their ability to operate in special environments without mechanical friction and lubrication, effectively avoiding wear and contamination. To overcome the complexity and maximum speed limitations, bearingless PMSMs (Permanent Magnet Synchronous Motors) have been developed. This study explains the electromagnetic driving principles of bearingless PMSMs based on generation conditions and the components of suspension forces. Furthermore, the permanent magnets arranged on the rotor are designed to alter the magnetic flux distribution using various magnetization patterns. FEA(Finite Element Analysis) was conducted for each model, and the derived results were compared and analyzed. This study investigates the performance of bearingless PMSMs based on different rotor permanent magnet arrangements, and draws a comprehensive conclusion.

https://doi.org/10.5370/KIEE.2025.74.4.747

이종석(JongSuk Lee) ; 김원종(WonJong Kim)

A real-time AC arc detection method was developed to predict potential fires and failures by detecting small arcs at an early stage, even on low-specification systems. To overcome the existing challenges related to limited data, time-series data closely resembling arc faults in real systems were acquired by introducing small arcs into the system. Data were collected from sensors installed inside switchgears and combined with simulation results to facilitate AI training, analysis, and validation. An AI framework for analyzing and predicting switchgear conditions was used to select appropriate AI algorithms and optimize input parameters. Additionally, an anomaly prediction device was developed to integrate precise time-series digital data acquisition and analysis for current and voltage, specifically for integrated AI switchgear. The developed AI-based real-time arc detection system successfully identified small arc phenomena through various filter chains and ensemble filters, demonstrating accurate detection performance.

https://doi.org/10.5370/KIEE.2025.74.4.753

조주희(Juhee Cho) ; 김수용(Suyong Kim) ; 이승재(Seungjae Lee) ; 김병택(Byungtaek Kim)

The current sensor of medium voltage DC converter station is basically required to have accuracy and linearity, and it must be immunity to the severe electromagnetic environment in the converter station. This paper analyzed the shielding structure and shielding material to satisfy the magnetic field immunity of MVDC-class current sensors. In order to evaluate the shielding characteristics against external magnetic fields, the test conditions specified in KS C 9610-4-8, 9 were referenced. The magnetic field strength and magnetic flux density distribution inside the sensor part, which can affect the current measurement, were calculated through three-dimensional finite element analysis. The analysis results are expected to be useful reference data for shielding design of fluxgate type current sensors