https://doi.org/10.5370/KIEE.2024.73.12.2165

김현태(Hyuntae Kim) ; 이후동(Hu-Dong Lee) ; 김용학(Yong-Hak Kim) ; 강석준(Seokjun Kang) ; 유덕기(Deokki You) ; 홍의준(Ryan EuyJun Hong) ; 강성현(Seong-Hyun Kang) ; 한창희(Changhee Han)

The research presents a stochastic optimization strategy for medium-voltage dc (MVDC) systems aimed at regulating power flows within distribution networks (DNs) under the coordination of a battery energy storage system (BESS). The study accounts for uncertainties arising from prediction errors in PV generators and load demand by employing a distributionally robust chance-constrained optimization approach. To cooperatively optimize the operation of MVDC systems and BESS in DNs, we stochastically relax reliability constraints on bus voltage, thereby maximizing the cost-effectiveness of the DNs. Numerical simulations demonstrate that by reducing the conservativeness of reliability constraints, the distribution system operator can achieve greater economic efficiency in electricity imports.

https://doi.org/10.5370/KIEE.2024.73.12.2175

이종욱(Jong-Uk Lee) ; 김영일(Young-Il Kim)

This study presents a method for calculating the maximum renewable energy capacity that can be connected to substations under a flexible interconnection policy. The method takes into account annual output control rates, transformer failure conditions, the proportion of controllable renewable facilities, and output control implementation rates. A case study conducted on substations in Jeollanamdo, where renewable capacity is near saturation, confirmed that higher output control rates and an increased proportion of controllable facilities lead to more connectable capacity.

https://doi.org/10.5370/KIEE.2024.73.12.2180

김영일(Young-Il Kim) ; 이종욱(Jong-Uk Lee) ; 노재구(Jae-Koo Noh) ; 김예리(Ye-Ri Kim)

To overcome the limitations of transformer acceptance capacity due to the rapid increase in renewable energy sources, the calculation method for the renewable energy connection acceptance capacity is changing from the previous method based on the installation capacity to the method based on the operating capacity. This paper studies a technology to predict the power generation of photovoltaic power plants connected to energy storage systems (ESS), which is necessary for predicting the operating capacity of transformer, using a multi-layer perceptron (MLP) based artificial neural network model.

https://doi.org/10.5370/KIEE.2024.73.12.2186

조건익(Gil-Dong Hong) ; 윤상윤(Il-Ji Me) ; 최준호(Soon-Sinl Lee) ; 정영범(Young-Sil Jang) ; 안선주()

Recently, distributed energy sources (DERs) have been increasingly integrated into distribution lines, raising concerns about over voltage issues. Accordingly, issues such as selection of installation locations and parameter settings for step voltage regulator (SVR) are emerging. This paper proposes an efficient method for determining the installation location and line drop compensator (LDC) settings based on load flow in a distribution line with multiple DERs. The proposed method consists of three steps: initial voltage profile analysis, SVR location selection, and LDC parameter setting. It aims to maintain the customers’ voltage close to the nominal value and mitigate voltage violations. The proposed method was verified through various case studies considering constraints in distribution line with multiple photovoltaic (PV) generators, and its effectiveness was proven by the results of case studies.

https://doi.org/10.5370/KIEE.2024.73.12.2196

박재현(Jae-Hyun Park) ; 박만기(Man-Ki Park) ; 권재호(Jea-Ho Kwon) ; 이종호(Jong-Ho Lee)

This case was occurred in February 2021 during the commissioning of the power plant. Through three checks were carried out, It was confirmed the cause of the current unbalance is the voltage unbalance, vice versa. The two faults were occurred five hours apart and the first fault occurred when a large motor was activated. It tripped as soon as it started and the startup transformer lost power. Second fault occurred when another large motor was activated. Six large motors were shut down at the same time. The check points and analysis results are as follows. In the process of switching the load of 11 busbar, the ground current of the ① ㎸ startup transformer decreases as the unbalanced current of the main transformer decreases. In the process of shutdown of a large ② motor, the rate of the current unbalance affected the rate of the voltage unbalance by 5.3 times. In the process of blackout of ③ the startup transformer, the cause of voltage unbalance is the current unbalance.

https://doi.org/10.5370/KIEE.2024.73.12.2203

윤성민(Seong-Min Yoon) ; (Peng Y. Lak) ; 남순열(Soon-Ryul Nam)

During the energization of transformers, the prevalent method for configuring the differential protection blocking logic involves setting it based on the second-harmonic content of magnetizing inrush. The protection IED can mal-operate as the second-harmonic content decreases during the magnetizing inrush in modern transformers. This paper presents a method for preventing the maloperation of differential protection by estimating the second-harmonic content using Digital Twin. In this study, the potential second-harmonic content from determined transformer conditions is estimated using Digital Twins. Comparative analyses under various conditions are provided through PSCAD simulations. Finally, based on the results of PSCAD simulations performed on transformers with different initial characteristics, a method is proposed to prevent maloperation of differential protection through the estimation of second-harmonic content.

https://doi.org/10.5370/KIEE.2024.73.12.2211

장홍재(Hong-Jae Jang) ; 김기찬(Ki-Chan Kim)

According to environmental policy, the vehicle is changed from an internal combustion engine car to an electric vehicle using an electric drive motor. A permanent magnet type is the main driving motor. Environmental pollution factors caused by the production of permanent magnets, i.e., rare earths, are causing problems that adversely affect the environment more than environmental pollution factors caused by internal combustion engine cars. Therefore, it is necessary to develop a WFSM (Wound Field Synchronous Motor) by requiring a dilute earth motor instead of a permanent magnet motor. The developed motor requires shape application to improve torque ripple, noise, and vibration because it is difficult to apply design elements such as rotor Skew. In this paper, the characteristics of applying balancing holes to rotor surfaces are compared to improve noise, vibration improvement, and motor operation safety due to ripple increase by field winding design for improving field inverter control

https://doi.org/10.5370/KIEE.2024.73.12.2217

손락원(Rak-Won Son) ; 이주(Ju Lee)

The industry is experiencing a gradual shift towards electrification, with various types of equipment replaced by electrical rotating machines. Also, the unit capacity of large machines is gradually increasing to improve the machine's efficiency and reduce operating costs. Modularization is essential to facilitate the assembly of these machines and repairs in case of failure. In this paper, the authors review the basic slot/pole combinations that allow modular winding and propose a 36-slot/10-pole fractional-slot concentrated winding with a single layer. This method could significantly improve the machines' fault-tolerance performance. The authors have conducted a thorough examination of the electrical performances of this proposed winding method, comparing it to a prototype shaft generator to provide a comprehensive analysis. The no-load test result shows the mutual consistency of the finite element analysis, reinforcing the confidence in the validity of this research

https://doi.org/10.5370/KIEE.2024.73.12.2224

김종겸(Jong-Gyeum Kim)

Induction motors can cause problems with the electrical quality of other devices voltage drop due to the high starting current when starting, so the reactor starting method is often used when the capacity is large. Since this induction motor consists of resistance and reactance components, the power factor is relatively low. When an induction motor starts, the reactive power is greater than the active power, so the power factor is low. As the speed increases, the sizes of the active and reactive power are reversed, and the power factor improves. In reality, the power factor when driving at rated speed is lower than that required by the power company, so the power factor is compensated with the help of a reactive power compensation device such as a capacitor. In the reactor starting method, capacitors are sometimes installed at the front and rear ends of the reactor. Reactor startup requires a long startup time and a long operating section with a low power factor, and the power characteristics appear different during startup depending on their installation location. In this study, we analyzed how changes in current and power affect the power factor during startup at the installation location of the capacitor in the reactor startup method.

https://doi.org/10.5370/KIEE.2024.73.12.2231

이찬기(Chan-Gi Lee) ; 이광운(Kwang-Woon Lee) ; 김상일(Sang-Il Kim)

To control grid-connected inverters, grid synchronization is essential, and a Phase-Locked Loop (PLL) technique is commonly applied for this purpose. Among various PLL techniques, the Second Order Generalized Integrator (SOGI)-PLL is widely used due to its robustness against external noise. However, when the grid voltage contains a DC offset, the SOGI-PLL exhibits steady-state error, which is a limitation. To address this issue, this paper applies the Cascade SOGI (CSOGI)-PLL, which has a robust characteristic against DC offsets. Applying the CSOGI-PLL requires designing the parameters of a fourth-order transfer function and tracking the grid frequency by adjusting the resonant frequency. In this paper, a simplified design method using a second-order system for the damping coefficient of the CSOGI with a complex transfer function is proposed. A separate PLL is employed to update the resonant frequency of the CSOGI in response to rapid frequency variations, and frequency limitation and anti-windup control are applied to ensure stability during transients. To validate the effectiveness of the proposed method, simulations and experiments were conducted.

https://doi.org/10.5370/KIEE.2024.73.12.2239

김민수(MinSu Kim) ; 이동희(Dong-Hee Lee)

This paper presents a improved sensorless open angle control scheme of the motorized louver system. In the proposed sensorless angle control method, the open angle of the louver blade can be controlled by the environmental temperature without any additional position sensor. In order to reduce the integrated angle error of the sensorless estimation, hysteresis open angle reference according to the temperature band is proposed. In the proposed control scheme, the open angle is not continuously determined. The 5-step open angle reference is determined by the 0.5 Celsius degree to reduce the low-speed operation region. Furthermore, the sensorless open angle controller, the non-linear reference voltage model is designed to consider the non-linear friction torque of the geared motor and louver blade. The proposed non-linear voltage reference model is designed to reduce the estimation error at the zero to low-speed region and overcome the non-linear friction torque. The integration of the voltage reference is reset to zero at the ±0.5 degree error region. In order to verify the proposed method, the proposed sensorless open angle control scheme of the motorized louver system is compared to the conventional sensorless control scheme. In the compared experimental results, the proposed method shows the reduced integrated snesorless angle error.

https://doi.org/10.5370/KIEE.2024.73.12.2247

홍성찬(Sung-Chan Hong) ; 정채림(Chae-Lim Jeong) ; 김국현(Kookyhun Kim) ; 박성주(Sung-Ju Park)

In a six-phase system, the difference in phase shift affects the winding factor and the degree of magnetic coupling, which influences the motor performance. However, the phase shift can lead to complex problems in motor design because both the phase shift and the rotor shape affect the performance of the motor. Therefore, the phase shift should be selected by considering its combination with shape parameters. This study analyzed the output characteristics according to the six-phase winding configuration and rotor geometry. The motor type used in this study is a synchronous reluctance motor, and the six-phase winding is classified into three types (Type 1, Type 2, Type 3) based on the phase shift. Afterwards, finite element analysis simulations were conducted by adjusting the values of the rotor shape parameters for each winding type. The result showed that the phase shift has a minimal impact on the rotor shape. Furthermore, analysis of the simulation confirmed that the parameters in the d-axis region are relatively more affected. Therefore, these analysis results can be utilized in future research related to the design of six-phase rotor shapes.

https://doi.org/10.5370/KIEE.2024.73.12.2255

김민성(Minseong Kim) ; 이주연(Juyeon Lee) ; 이준석(June-Seok Lee)

This paper proposes an improved CB-DPWM(Carrier-Based Discontinuous Pulse Width Modulation) method for the Vienna rectifier. The conventional CB-DPWM methods clamp the reference voltage to zero during intervals where the Vienna rectifier's normal operating condition is not met due to the offset voltage for DPWM operation. However, these additional clamping intervals cause an increase in the phase current THD(Total Harmonic Distortion). In this paper, an improved CB-DPWM method is proposed to reduce the phase current THD of the Vienna rectifier by minimizing intervals where the reference voltage is zero. Furthermore, neutral point voltage control can be applied in the non-clamping regions to mitigate fluctuations in the neutral point voltage. Consequently, the proposed method can enhance the phase current THD while maintaining the advantage of reducing switching losses. The effectiveness of the proposed method is verified through simulations and experimental results.

https://doi.org/10.5370/KIEE.2024.73.12.2264

김민솔(Min-Sol Kim) ; 고영종(Youngjong Ko)

The Cascaded H-Bridge (CHB) converter offers advantages in reliability, power quality, and efficiency over non-modular topologies. However, variations in electrical and thermal parameters within individual cells can lead to uneven lifetimes, causing disproportionate degradation of capacitors and power semiconductors. This necessitates individual lifetime management to enhance system reliability and lower maintenance costs. Active thermal management has been explored to reduce thermal stresses and improve power semiconductor reliability, yet the reliability of DC-link capacitors is often overlooked. This paper examines the impact of different modulation schemes on the reliability of DC-link capacitors and power semiconductors in CHB converters. It then proposes an optimal modulation strategy for managing component reliability and validates it through experiments.

https://doi.org/10.5370/KIEE.2024.73.12.2272

박형욱(Hyeong-Wook Park) ; 이재형(Jae-Hyung Lee) ; 윤재중(Jae-Jung Yun)

A state of charge (SOC) estimation method based on coulomb counting requires an accurate initial SOC value to reset the accumulated errors during the integration process. However, accurate initial SOC estimation is challenging for lithium iron phosphate (LFP) batteries due to their flat SOC-OCV curve and hysteresis characteristics. To address these problems, this paper proposes an enhanced coulomb counting based SOC estimation algorithm that incorporates hysteresis characteristics. The hysteresis characteristics of the LFP battery are analyzed through major loop and minor loop experiments. In order to determine the accurate initial SOC value that matches the measured open circuit voltage (OCV) in the hysteresis characteristics of the battery, the proposed method creates a new SOC-OCV curve consisting of the average curves of the major loop and charge/discharge curves of minor loops. The operating point (SOC, OCV) of the battery is determined by linear interpolation after selecting a curve according to the previous SOC and the state of the battery. A test bench is constructed to verify the SOC estimation performance of the proposed method. As a result of the experiment, the proposed method has a smaller SOC estimation error by reflecting the hysteresis characteristics of SOC-OCV more accurately compared to the existing improved coulomb counting method. The mean squared error (MAE) and root mean squared error (RMSE) of the SOC estimation of the proposed method are both less than 0.5.

https://doi.org/10.5370/KIEE.2024.73.12.2281

나석진(Seok-Jin Na) ; 이한민(Han-Min Lee) ; 김재원(Jae-Won Kim) ; 조인호(In-Ho Cho)

Conventional Slave-Master structured Battery Management Systems(BMS) face limitations due to the inherent constraints of their wired architecture. Also, the aging characteristics of propulsion lithium-ion batteries, which require management, vary depending on the operating environment and output conditions. When these limitations occur simultaneously, conventional BMS monitoring and control functions become inadequate for accurate fire risk prediction and effective safety management. So, we propose a Masterless BMS to overcome these limitations. The proposed BMS efficiently processes large volumes of battery data through wireless communication and enables sophisticated battery management through Artificial Intelligence(AI)-based battery state estimation using an external server. This paper analyzes efficient operational methods and optimal design approaches for applying a Masterless BMS architecture to propulsion Energy Storage Systems(ESS) for railway vehicles..

https://doi.org/10.5370/KIEE.2024.73.12.2289

김동환(Dong Hwan Kim) ; 고정수(Jeong Soo Ko) ; 김도현(Do Hyeon Kim) ; 임종헌(Jong-Hun Lim) ; 임제영(Je Yeong Lim) ; 이재인(Jaein Lee) ; 이병국(Byoung Kuk Lee)

This paper proposes a method for estimating the State of Charge (SOC) of lithium iron phosphate (LFP) batteries based on an experimental thermal model that considers internal battery temperatures. The proposed algorithm involves selecting a thermal model and suggesting methods for extracting parameters from the chosen model. Using the proposed thermal model, the algorithm compensates for temperature differences in cold and hot environments, thereby correcting the internal temperature to estimate SOC accurately. The effectiveness of the proposed algorithm is verified by estimating the SOC of electric vehicle LFP batteries and comparing the accuracy before and after correction.

https://doi.org/10.5370/KIEE.2024.73.12.2297

차주홍(Ju-Hong Cha)

Microwave plasma sources are used in nitride thin film deposition owing to their superior thin film characteristics and low processing temperatures. Slot antenna optimization was examined using finite element electromagnetic simulations to enhance power transfer efficiency and spatial uniformity related to individual slot propagation in microwave plasma sources. The propagation efficiency was measured based on the structure and position of the slot antenna by configuring the slot antennas in various orientations, including transverse, longitudinal, waveguide-rotated longitudinal, and hybrid slots. For the four types of slot positions, the lengths of the major and minor axes were varied from 5 to 60 mm and 2 to 10 mm, respectively. The increase in electromagnetic wave transmission through the slots and the power transfer uniformity according to individual slots depended on the major axis length, showing optimization at 50 55 mm.

https://doi.org/10.5370/KIEE.2024.73.12.2305

김성훈(Seong-Hun Kim) ; 김진규(Jin-Gyu Kim)

A nanosecond pulse dielectric barrier discharge (NPDBD) technique is attracting attention due to the fact that it produces uniform discharge and energetic electrons at atmospheric pressure. However, there has been a lack of research on NPDBD characteristics when high-repetition pulses can be processed quickly in a short period of time. Therefore, in this study, in order to identify the characteristics of high-repetition NPDBD according to flow rate, discharge characteristics were analyzed through discharge current and the Lissajous curve. As a result, as the flow rate increased, charge transfer decreased, wall voltage dropped due to a reduction in residual charge, and a current delay occurred. In addition, the breakdown voltage increased, and fillamentary discharge formed. This study's findings are expected to be extremely useful in the future when high-repetition nanosecond pulse power supply is used for efficient processing in a short period of time in a variety of plasma-based applications.

https://doi.org/10.5370/KIEE.2024.73.12.2312

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

This paper proposes a respiratory detection and localization system using a Phased Array Antenna (PAA). In this system, we used 4×4 phased array antenna with Left-Handed Transmission Line (LHTL) type phase shifter, which has advantages on miniaturization and simple control. The whole scanning area was 3 m × 3 m and the area was divided into 4 squares whose individual area was 1.5 m × 1.5 m. Beam steering of the PAA was conducted for four directions (up down left right) from the center with +30°, -30° bias angle to scan these individual areas. Using the fabricated PAA and respiration detection algorithm, this radar system was able to locate the position of human respiration and its frequency with 94 % accuracy from 3.5 m distance.

https://doi.org/10.5370/KIEE.2024.73.12.2318

서성부(Seongbu Seong) ; 김우곤(Woogon Kim) ; 서예준(Yejune Seo) ; 박도현(Dohyun Park) ; 김형종(Hyoungjong Kim) ; 이권희(Kwonhee Lee) ; 이호섭(Hosub Lee) ; 강승택(Sungtek Kahng)

Satellite communication requires a long-distance wireless link, which necessitates a sufficiently high gain from the antenna, the entry point of a wireless communication device. Array antennas, commonly used to achieve high gain, face challenges such as cumulative metal and dielectric losses in the feed system and large physical volume. The use of optical curved lenses for high gain at millimeter-wave frequencies has been proposed, but such lenses have drawbacks, including large volume, as well as high manufacturing, installation, and maintenance costs. This paper addresses the design of a planar metasurface lens capable of amplifying millimeter-wave signal strength, serving as an alternative to array antennas and curved lenses. A planar lens with a 100 mm × 100 mm area that enhances the radiation gain of K-band antennas, a representative frequency range for satellite communication, by more than 9 dB is implemented. To verify the design, the radiation characteristics of the source antenna combined with the metasurface lens are measured and analyzed using a Vector Network Analyzer (VNA) and an Anechoic Chamber. Observations of the proposed antenna's frequency response characteristics demonstrate that the antenna gain increases by more than 10 dB in the K-band, and the beamwidth narrows, indicating high directivity.

https://doi.org/10.5370/KIEE.2024.73.12.2326

이수빈(Su-bin Lee) ; 박재범(Jae-beom Park) ; 조현종(Hyun-chong Cho)

As the number of livestock farms in Korea decreases, the scale of individual farms has grown, necessitating advanced methods for efficient livestock management. Recent studies have increasingly applied computer vision and deep learning to monitor livestock behavior, health, and disease, aiming to enhance productivity and animal welfare. This study introduces a real-time cattle behavior detection system utilizing YOLOv9, a state-of-the-art object detection algorithm known for its high accuracy and efficiency. The YOLOv9-s model, optimized for real-time detection with minimal computational overhead, was employed. To further enhance detection accuracy, we incorporated AutoAugment, a data augmentation technique that automatically selects the optimal augmentation policies, and mixup, a method that improves model generalization by creating new training samples through linear interpolation of existing ones. The application of AutoAugment improved the mean Average Precision (mAP) from 0.926 to 0.941, while the addition of mixup raised the mAP to 0.947, representing a 2.1% performance increase. These results confirm the system's capability for accurate and efficient real-time detection of cattle behavior.

https://doi.org/10.5370/KIEE.2024.73.12.2333

김광현(Gwang-Hyeon Kim) ; 오하령(Ha-Ryoung Oh) ; 성영락(Young-Rak Seong)

Recently, Transformer has demonstrated excellent performance by effectively addressing the long-term dependency issues of traditional time series prediction models. However, Transformer also has limitations, such as a lack of ability to learn the sequential characteristics of time series data and the need for significant computational resources due to their complex structure. This paper proposes an EMformer model based on the Transformer architecture to overcome these limitations and improve prediction accuracy. EMformer reduces computational costs by replacing the decoder with an MLP and enhances sequential feature extraction by replacing the feed-forward neural network in the encoder with depthwise separable convolution. The model's performance is evaluated using power consumption dataset and compared with other time series prediction models. The results show that EMformer improve performance by up to 57.3% in MAPE and 30.75% in RMSE compared to other models.

https://doi.org/10.5370/KIEE.2024.73.12.2341

문성욱(Seonguk Moon) ; 조영완(Youngwan Cho)

This paper proposes a two-joint arm model using reinforcement learning in an air hockey simulation environment. Reinforcement learning, a method where an agent interacts with the environment to learn optimal behavior strategies, is applied to a physics-based air hockey simulation in this study. Air hockey is a game with simple rules where the puck must be directed into the opponent's goal area. Unlike conventional simple control algorithms, this study aims to implement diverse actions and varied racket trajectories through reinforcement learning. The focus of this paper is on the implementation and performance evaluation of a two-joint arm model that primarily learns a defensive play style. The two-joint arm model is assessed for its ability to respond to various situations and its basic control performance. Through this evaluation, the study aims to verify the fundamental potential of reinforcement learning-based agents and suggest future research directions.

https://doi.org/10.5370/KIEE.2024.73.12.2350

서기성(Kisung Seo)

Optimization of multi-layer, multi-element systems such as deep learning is an NP-hard problem that requires determining the number of layers, the number of elements in a layer, and the types of elements. In such a system, deleting redundant elements to reduce the size while maintaining performance is crucial to conserve resources and improve the efficiency of the system. This is a very complex and challenging problem because it consists of a large number of multi-layered and multi-element systems with a huge search space. Evolutionary computation is widely used for large-scale optimization problems due to its high efficiency, but it is difficult to apply due to the characteristics of evolutionary computation when the calculation of the fitness function is complex. To solve this problem, we propose a technique that dramatically reduces the search space by improving the representation of the gene. We verify its feasibility by applying it to a case study, CNN pruning. We use the ResNet56 model for the CIFAR10 dataset and compare it with existing pruning approaches.

https://doi.org/10.5370/KIEE.2024.73.12.2355

전병주(Byoung Ju Jeon) ; 김동헌(Dong Hun Kim)

In this paper, we propose an AI inspection system based on deep learning to analyze the product image of the action loader (linear actuator system) inside an automatic car door handle. The proposed system aims to classify the model of the action loader and detect its defects. In addition, the paper compares performance evaluation using CNN(Convolutionary Neural Networks)-based Xception and Inception models, respectively. Due to low defective image data, data augmentation and transfer learning techniques are applied to improve the performance. As an experimental result, Xception showed comparatively better results than Inception. These experimental results were confirmed using Accuracy, Precision, Recall, and F1-Score based on the Confusion Matrix performance indicator. The proposed CNN-based inspection system presents practical applicability in the field of automobile part classification and defect detection.

https://doi.org/10.5370/KIEE.2024.73.12.2363

하창완(ChangWan Ha) ; 장은성(Eunseong Jang) ; 최정민(Jeongmin Choi) ; 장현배(Hyunbae Chang) ; 조형기(HyungGi Jo)

This research proposes a novel method for determining the location of vessels using a multi-sensor complex consisting of sound, magnetic, and depth sensors. The goal is to create a reliable and precise system that can overcome the limitations of individual sensors by integrating their data. Gaussian Process Regression (GPR) is employed to interpolate sparse data collected by the multi-sensor system, while a deep learning model based on the Transformer architecture is used to estimate the vessel's position. The system is designed to enhance accuracy and robustness, particularly in noisy marine environments. Our method shows potential for real-time applications in underwater localization, particularly in areas with high noise and interference.

https://doi.org/10.5370/KIEE.2024.73.12.2371

한병찬(Byeong-Chan Han) ; 강민제(Min-Jae Kang) ; 송성호(Seong-Ho Song) ; 김호찬(Ho-Chan Kim)

This paper proposes a memory transformer Q-learning network(MTQN) algorithm to improve existing deep reinforcement learning algorithms. MTQN is configured by combining transformers with existing deep reinforcement learning models to model sequence systems more efficiently, and the gating mechanism of LSTM is additionally used for using the transformer. The proposed algorithm is compared and analyzed with DQN, a representative deep reinforcement learning algorithm, and its modified algorithms, targeting cart-pole system, a representative reinforcement learning benchmark environment. The simulation extracts and compares the evaluation score, cart position, and pole angle of cart-pole system, and shows that the proposed algorithm learns the fastest and most stably.

https://doi.org/10.5370/KIEE.2024.73.12.2381

김성주(Sung-Ju Kim) ; 백명기(Myung-Ki Baek) ; 곽창섭(Chang-Seob Kwak) ; 성민제(Min-Je Sung) ; 강원태(Won-Tae Kang) ; 허영근(Young-Gun Heo) ; 김재형(Jae-Hyung Kim)

This study applies simulation-based digital twin technology to a centrifugal pump system for real-time monitoring and performance prediction. The analysis method is to analyze the flow field of the centrifugal pump system by flow analysis, and the structural analysis of the impeller is performed using the analyzed flow field. 100 DPs(Design Points) were selected using the Latin Hypercube Sampling(LHS) method, and Reduced Order Models(ROM) were created by learning 90% of the DPs. The RS ROM(Response Surface ROM) predicts flow and pressure difference(PD), and the Static ROM predicts pressure, speed, and stress on the impeller. We conducted a lift and efficiency comparison experiment, and the results showed an error of 0.3 [m] in lift and about 10 [%] in efficiency. Pressure difference and flow comparison experiments showed an average pressure difference error of 0.927 [kPa] and an average flow error of 20.00 [liter/min]. The virtual sensor predicted the speed, pressure, and stress on the impeller in the event of cavitation.

https://doi.org/10.5370/KIEE.2024.73.12.2391

김지수(Ji-Soo Kim) ; 정경용(Kyungyong Chung)

In this paper, we propose an algorithm that combines the YOLO model with edge detection techniques to effectively detect cracks occurring in structures. Existing manual inspection methods are time-consuming, costly, and suffer from poor accuracy. To solve this problem, this study proposes an automated deep learning-based detection method, specifically using a YOLO model and a canny edge detection algorithm to perform more precise crack boundary extraction. The proposed algorithm consists of two steps. The first step is image preprocessing, which uses several preprocessing techniques such as gray-scale conversion, median blur, normalization, and brightness inversion to highlight the edges of cracks. In the second step, the canny edge detection algorithm is used to detect the boundaries of the cracks in the preprocessed image, and the YOLOv8 model is used to learn and detect the location of the corresponding cracks. The results show that the best performance is achieved when the YOLO model is trained using 6 steps of preprocessing and canny edge detection.

https://doi.org/10.5370/KIEE.2024.73.12.2398

유혜정(Hyejeong Ryu) ; 최진우(Jinwoo Choi)

An efficient local kernel-based Gaussian Process (GP) grid map approximation method is proposed for mobile robot exploration using 2D Light Detection and Ranging (LiDAR) sensors. While the conventional GP grid map captures structural correlations, its prediction performance is affected by the training data from LiDAR measurements, and the computational burden increases with the number of training and test points. To reduce computational complexity and improve prediction consistency, we construct a local kernel using trained hyperparameters and perform convolution operations with a modified local grid map derived from LiDAR measurements. Experiments using LiDAR measurements obtained from a robot navigating an indoor environment verified that the method efficiently computes the GP grid map and frontier probability, making it suitable for practical frontier-based exploration.

https://doi.org/10.5370/KIEE.2024.73.12.2407

이한상(Hansang Lee) ; 김도혁(Dohyuk Kim) ; 김형철(Hyungchul Kimg)

As the integration of active equipment based on power electronics technology is expected to form an eco-friendly railway system, it is necessary to review the durability of the AC railway system against harmonic currents injected from converters. Recently, a dissipation of lightning arrester had been occurred due to overvoltage caused by harmonic resonance during the operation of railway vehicles, therefore the need for system analysis to avoid resonance has increased. This paper defined the length of the catenary system and the capacity of the Scott transformer as factors affecting the resonance harmonic order and magnitude of impedance in AC railway systems, and analyzed their effects. In addition, the change in the resonance harmonic order and the magnitude of impedance whether RC bank is applied or not was analyzed.

https://doi.org/10.5370/KIEE.2024.73.12.2414

신승권(Seung-Kwon Shin) ; 송한솔(Han-Sol Song)

This study proposes an evaluation methodology that considers cooling and heating degree days(CDD, HDD) to assess the energy use efficiency(DDEER) of major railway stations. Although railway stations have been classified as transportation and logistics facilities, and thus exempt from energy efficiency regulations, they are now included under the carbon emission control measures of the 2050 Carbon Neutral Roadmap. Consequently, systematic energy management is required for railway stations. To address this need, this study developed a methodology to evaluate energy use efficiency by comparing energy consumption with cooling and heating degree days using a standardized Z-score. Data were collected from 28 major KTX stations across the country over a three-year period (2021 2023), including both energy consumption and meteorological data. The analysis was then applied to three selected stations ? ? Yongsan, Suwon, and Yeongdeungpo to assess their energy efficiency. The results indicate that energy efficiency is generally lower ? during the heating season compared to the cooling season, and significant fluctuations in energy use were observed during transitional months (March and October). These findings suggest a need to review and improve the internal operation standards for heating and cooling systems. The proposed methodology provides a valuable tool for prioritizing energy management and improving energy efficiency in railway stations, contributing to the overall goal of achieving Zero Energy Building (ZEB) certification.

https://doi.org/10.5370/KIEE.2024.73.12.2422

오세화(Sehwa Oh) ; 김성철(Seong Cheol Kim) ; 박철민(Chulmin Park) ; 박영(Young Park)

As the construction of underground railways for 180 km/h-class metropolitan express trains increases in Korea, ensuring stable current collection performance in tunnels becomes essential. In tunnel sections, the reduced separation between the pantograph and the overhead contact line (OCL) increases the risk of arc occurrences and affects pantograph dynamics. This study analyzes technical standards for verifying current collection performance between the pantograph and the OCL in underground sections at speeds up to 180 km/h. It was found that conventional video monitoring at 30 fps cannot detect arcs of 5 ms duration at 170 km/h because the train moves about 47 mm during a 1 ms exposure. Detecting such arcs requires a camera with 400 fps or higher. Additionally, with a camera resolution of 1936 × 1216 pixels and a viewing area of 1200 × 800 mm, the minimum detectable defect size is approximately 0.62 mm horizontally and 0.66 mm vertically. The proposed methods improve current collection performance verification, contributing to safer and more efficient underground railway operations.

https://doi.org/10.5370/KIEE.2024.73.12.2427

김영수(Young Soo Kim) ; 서기범(Ki Bum Suh)

Thermal runaway is the representative cause of lithium-based storage batteries currently operating in Korea's railway electricity sector, and a battery module pack with a battery cell surface temperature detection function suggested as an effective variable that can detect thermal runaway abnormalities of lithium batteries in advance and a card-type battery-type product that can prevent safety accidents caused by heavy object drop accidents were developed and compared. Existing products have a single mechanism that operates the protection circuit at 75°C by detecting only the temperature inside the battery module. This development was carried out in a two-stage temperature monitoring method of module pack and cell unit sensing. In addition, an alarm is generated when the temperature is 55°C so that users can recognize and safely respond to dangerous situations before a fire accident and before reaching 65°C, the recommended maximum temperature for safety standards.

https://doi.org/10.5370/KIEE.2024.73.12.2435

김윤호(Yunho Kim) ; 이정환(Jung-hwan Lee) ; 김문수(Moon Soo Kim) ; 윤예현(Ye Hyeon Yoon) ; 김영철(Young Cheol Kim)

Aluminum alloys used in part manufacturing with LPBF, particularly the Al-Si-Mg series, are widely utilized across various industries due to their excellent thermal conductivity and corrosion resistance. However, the high cost of these materials is a significant drawback, as high-purity metal powder is required to enhance dimensional accuracy and surface roughness. Therefore, powder reuse is essential. This study investigates the effects of powder reuse on the mechanical properties of AlSi10Mg alloy produced LPBF. The powder was analyzed over 16 times, including the initial use of new powder and 15 times reuse. Parts were manufactured in 15 times, and the changes in the properties of both the powder and the parts were analyzed. The results indicate that the powder can be reused up to 15 times without significant issues.

https://doi.org/10.5370/KIEE.2024.73.12.2439

김윤호(Yunho Kim) ; 최성민(Sung-Min Choi) ; 김영철(Young Cheol Kim) ; 송준부(Song Joon Boo) ; 홍민호(Min-Ho Hong)

As the prevalence of partial tooth loss in elderly patients increases, accurate RPD frameworks becomes crucial for comfort and oral health. This study compared the dimensional accuracy of RPD frameworks fabricated at various build angles (0°, 15°, 30°, 45°, and 60°) using Additive Manufacturing (AM) technology. The samples were made from CoCr28Mo6 alloy powder. The accuracy of each framework was evaluated using root mean square (RMS) values and color deviation maps. The results showed that as the build angle increased, the accuracy decreased, especially in critical areas like clasps. One-way ANOVA analysis revealed statistically significant differences between groups (p < 0.001), with Tukey’s post-hoc tests indicating substantial discrepancies between build angles greater than 15°. These findings demonstrate that the build angle in AM significantly affects the precision of RPD frameworks, highlighting the need for optimized angle selection in clinical applications.

https://doi.org/10.5370/KIEE.2024.73.12.2444

최진욱(Jin-Wook Choe) ; 이춘권(Chun-Kwon Lee)

This paper proposes a optimization method for estimating the conductance of cable insulation, a parameter closely related to the dielectric losses and overall health of the cable. The proposed method utilizes a stepped frequency waveform reflectometry technique, which injects a frequency range of sinusoidal signals into the cable and analyzes the reflected signals to assess the attenuation and phase shift caused by the cable's physical and electrical properties. By optimizing the conductance based on the measured reflection coefficients and applying the Levenberg-Marquardt algorithm to minimize errors, the method accurately estimates the internal conductance of the cable's insulation layer. The ability to quantitatively assess and monitor conductance over time provides a valuable tool for early detection of insulation degradation and prediction of remaining cable life.

https://doi.org/10.5370/KIEE.2024.73.12.2449

최용은(Yong-Eun Choi) ; 오효석(Hyo-Seok Oh) ; 박상녕(Sang-Nyeong Park) ; 고준영(Joon-Young Ko) ; 김재문(Jae-Moon Kim)

In the DC link of a lithium-ion battery system, common mode voltage(CMV) is generated due to grounding. CMV is mainly generated by grounding imbalance, ground potential difference, and signal interference within the system, and affects the performance of the battery system. In particular, CMV can cause electromagnetic interference(EMI) in the high-frequency range or reduce efficiency during the power conversion process, which can adversely affect the reliability of the power system. Therefore, in this paper, we analyzed the characteristics of CMV generated due to grounding of a lithium-ion battery through hardware experiments under various conditions. To this end, we compared and analyzed the CMV characteristics generated during charging and discharging according to the ESS operation during the day and night. In addition, we analyzed the variation of CMV according to the presence or absence of load when charging the battery through photovoltaics(PV) and when connected to the grid.

https://doi.org/10.5370/KIEE.2024.73.12.2456

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

This study presents a methodology for optimizing asset management of power equipment by quantitatively evaluating their condition. The approach enables the development of effective maintenance strategies, ensuring both operational reliability and economic efficiency. Existing frameworks like the UK’s CNAIM provide a robust model for power equipment assessment but are limited in direct application to South Korea due to data and environmental differences. This study proposes a health index calculation method tailored to the Korean context. By focusing on key deterioration factors and integrating local operational data, the method enables accurate condition assessment and informed maintenance prioritization. It is adaptable to environments with limited failure records or operational data. The proposed method is expected to extend equipment lifespan, reduce costs, and enhance asset management.

https://doi.org/10.5370/KIEE.2024.73.12.2463

백종섭(Jongseb Baeck) ; 황현주(Hyun-Joo Hwang) ; 남기준(Gi-Joon Nam) ; 이인호(In-Ho Lee)

Fire-resistant cable in building refers to the cable that can withstand flames and maintain electrical or communication transmission functions for a certain period of time in the event of a fire. This plays an important role in minimizing damage to human life and properties by keeping critical systems running. For power systems, regulations for cables have been revised to KS C IEC 60331-1,2 to meet international standards for high fire-resistant performance at the low voltage (LV) level. However, further development and standardization of high fire-resistant medium voltage (MV) power cables are needed. In case of communication systems, the use of flame retardant cables is still allowed in Korea, so the development and standardization of fire-resistant communication cables are needed. High fire-resistant medium-voltage and telecommunication cables are developed, and the results are described in this paper. It is believed that legislation and practicalization of fire-resistant cables for power and communication cables are essential to minimize damage to human life in the event of a fire and, consequently, to realize a safer society.

https://doi.org/10.5370/KIEE.2024.73.12.2469

백창용(Chang-Yong Baek) ; 이동명(Dong-Myung Lee) ; 강태은(Tae-Eun Kang)

The need to achieve carbon neutrality and the acceleration of urbanization are increasingly prominent. The demand for apartment complexes is rising, leading to an emerging need for increased electrical facility capacity. Although various attempts have been made to calculate the electricity requirements for each apartment unit, the improvements have not yielded significant results. Therefore, this paper proposes a method for calculating the electrical capacity for each apartment household unit, taking into account actual electricity usage conditions. It addresses the challenges in calculating the electrical capacity of domestic apartment complexes through design cases in major cities in Gyeonggi province. This paper describes the key issues related to the calculation of household electricity consumption in apartment complexes in South Korea and compares and analyzes the electricity calculation methods between OECD countries and domestic practices

https://doi.org/10.5370/KIEE.2024.73.12.2476

김동민(Dongmin Kim) ; 서정진(Joungjin Seo) ; 차한주(Hanju Cha)

This paper analyzes the magnitude of leakage current due to parasitic capacitance and cable impedance generated by structural characteristics in Battery energy storage systems(BESS). The cable was measured using an LCR meter and employed in simulations and experiments. A leakage current equivalent model was constructed and validated through comparison with experimental data. In the case with no cable impedance, the leakage current path in the IT grounding method includes paths through the LCL filter of the PCS, transformer impedance, and parasitic capacitance. However, when cable impedance is added, it was confirmed that the leakage current magnitude increases because an additional leakage current path through the cable is introduced. Therefore, the magnitude of the leakage current was 43.3 mArms before the cable was added, but after adding the cable impedance, it increased to 159.8 mArms, demonstrating a 3.7-fold increase.

https://doi.org/10.5370/KIEE.2024.73.12.2482

황인기(Inki Hwang) ; 유광열(Kwangyeol Yoo) ; 김명진(Myungchin Kim)

In this study, the effectiveness and performance of machine learning approaches for predicting surface flashover voltage of thermally stressed specimen were studied. In particular, k-NN (k-Nearest Neighbors) and ANN (Artificial Neural Network) technique were considered. Data from flashover experiments and features of the electrical field distribution were used to train the machine learning models. The overall process of implementing the prediction model is introduced. While both approaches were optimized through hyperparameter tuning, it was shown that the ANN approach shows slightly better prediction performance. In particular, ANN seemed to show satisfactory performance for specimen that experienced higher stress conditions. Although the k-NN approach showed some limitations compared to the ANN approach, the k-NN approach could also show potential for applications thanks to its rather simple approach and requiring less computation resources.

https://doi.org/10.5370/KIEE.2024.73.12.2490

임용배(Young-Bae Lim) ; 정다연(Da-Yeon Jung) ; 김창수(Chang-Soo Kim) ; 신동준(Dong-Joon Shin)

This paper utilized statistical data to assess the losses caused by electrical fires and proposed a risk assessment methodology grounded in electrical fire statistics. The proposed methodology quantified the monetary value of human and material losses to evaluate the economic benefits of electrical fire prevention measures. These findings provide a foundation for assessing the cost-effectiveness of electrical fire prevention strategies, including the implementation of technologies such as arc-fault circuit interrupters and other advanced prevention devices. The study compared the time in service of a building's electrical wiring with the frequency of electrical fires and confirmed that wiring aging, previously regarded as a major hazard, is less significant than earlier assessments suggested. Furthermore, an analysis of electrical fire distribution across various types of multi-use facilities revealed how fire frequency and severity vary, offering insights into targeted strategies for mitigating electrical fire risks.

https://doi.org/10.5370/KIEE.2024.73.12.2496

태동현(Dong-Hyun Tae) ; 김유나(You-Na Kim) ; 모영규(Young-Kyu Mo) ; 황민(Min Hwang) ; 우필성(Pil-Sung Woo) ; 김영석(Young-Seok Kim) ; 송길목(Kil-Mok Shong)

Recent fire incidents in large-scale ESS(energy storage system) have raised concerns about the electrical risks inherent in their operation. Specifically, the CMV(common mode voltage) generated by the PCS(power conversion system) in ESS can impose continuous electrical stress on the battery and DC circuits, potentially leading to severe issues such as battery insulation failure and increased electrical noise. CMV primarily arises from interactions between switching components and parasitic capacitance, and this phenomenon becomes more pronounced in large-scale ESS, accelerating battery degradation and increasing the risk of fire hazards. This study analyzes the mechanisms of CMV generation in ESS and quantitatively evaluates CMV variations during charging and discharging processes through linear regression analysis. By modeling the relationship between ESS output power and CMV, system degradation and potential risks are predicted by examining CMV fluctuations during operation. Experimental results indicate that as charging power increases, the CMV magnitude also rises, intensifying electrical stress on the battery and DC circuits, while CMV decreases during discharging. These findings suggest that CMV trends can serve as indicators for early detection of system abnormalities by comparing them with trends observed in healthy systems. This approach has the potential to enhance ESS safety and help prevent incidents such as fires by identifying potential risks in advance.

https://doi.org/10.5370/KIEE.2024.73.12.2507

김지연(Ji-Yeon Kim) ; 최상재(Sang-Jae Choi) ; 전주현(Ju-Hyeon Jeon) ; 이건호(Geon-Ho Lee) ; 송길목(Kil-Mok Shong)

Aligned with global endeavors to address environmental challenges such as aging power infrastructures, transition from fossil fuels, and combatting greenhouse gas emissions and climate change, countries worldwide are actively pursuing energy transition policies. Consequently, energy storage technology emerges as a critical necessity, driving diverse research efforts to ensure the safety and efficacy of Energy Storage Systems components, including Power Conversion Systems (PCS), batteries, protective mechanisms, control systems, and installation protocols. This paper presents an analysis of property variations in separators, pivotal in insulating and facilitating lithium-ion reactions, employed in commercially available energy storage systems(ESS, UPS etc.) featuring NCM series batteries in Korea. Leveraging Battery Management System (BMS) data, this study forecasts battery characteristic shifts and conducts physical disassembly to validate internal alterations. Extracted samples of separator membranes from fresh cells, operational cells without anomalies, and cells exhibiting irregularities are subjected to electrochemical measurements, presenting observed changes.

https://doi.org/10.5370/KIEE.2024.73.12.2514

전주현(Juhyeon Jeon) ; 이건호(Gun-Ho Lee) ; 김지연(Ji-Yeon Kim) ; 최상재(Sang-Jae Choi) ; 송길목(Kil-Mok Shong)

In this paper, methods for diagnosing degradation using ESS voltage data were analyzed. With the domestic and international market size expected to grow, the aim is to diagnose battery degradation before a fire occurs to address the fundamental issues. Voltage data, which can be applied as common data to all ESS sites, was used for statistical analysis, specifically the Interquartile range method and the Z-score method. Analysis of dozens of normal and problematic sites classified them with about 50% accuracy, but since the number of normal sites was much higher, the actual accuracy is presumed to be lower. Feeling the limitations of statistical analysis, additional analysis is being conducted using a deep learning Autoencoder model. Although data processing and preprocessing techniques are still lacking, improvements are being made to prevent ESS fires in advance.

https://doi.org/10.5370/KIEE.2024.73.12.2521

김채원(Chae Won Kim) ; 김용희(Yong Hee Kim) ; 이세희(Se-Hee Lee Soo Ko)

We explored real-time internal resistance (rt-IR) and its trends as a critical diagnostic parameter for monitoring the safety of lithium-ion batteries. With growing demand for battery devices, repeated charge and discharge cycles accelerate performance degradation, leading to potential safety concerns. By monitoring rt-IR, this research aims to detect early signs of battery degradation. The rt-IR was evaluated by testing several time intervals using current and voltage data under constant current state-of-charge (CC-SoC), obtained from the Oxford battery degradation dataset. Our results showed that the rt-IR values were relatively higher at lower SoC level, particularly below 30% of CC-SoC, indicating elevated safety risks. Based on these results, we proposed that rt-IR and its trends can be an important indicator for diagnosing lithium-ion battery safety. This real-time monitoring approach with rt-IR provides valuable insights into maintaining safety and efficient battery operation in energy storage systems (ESS).

https://doi.org/10.5370/KIEE.2024.73.12.2528

김유나(You-Na Kim) ; 태동현(Dong-Hyun Tae) ; 모영규(Young-Kyu Mo) ; 황민(Min Hwang) ; 우필성(Pil-Sung Woo) ; 김영석(Young-Seok Kim) ; 송길목(Kil-Mok Shong)

Recently, the industry has been paying attention to energy storage devices to resolve intermittency as renewable energy generation increases. However, the ESS (Energy Storage System) industry continues to shrink due to a series of ESS fires that have occurred recently in Korea. Accordingly, KESCO (Korea Electrical Safety Corporation) analyzed 59 cases of domestic energy storage device fires or failures. As a result, the battery interior, battery room environment, and operation were defined as being very important to safety. This paper presents a digital twin framework that performs real-time monitoring and simulation to perform various failure cause analysis and safety diagnosis. This digital twin framework presents about 21 data management items, including 8 items for protection maps, 6 items for artificial intelligence algorithms, and 3 items for battery status, 3 items for battery environment, and 1 item for battery operation related to data collection.