Byeong-Woo Kim

Prof. Byeong-Woo Kim


Research Focus & Metrics

Academic Career Length

16Years (2011-Present)

(Calculated once Scilit publication data is synced: current year − year of first publication + 1)

Representative Publications

View All Publications
29 April 2020 in Sensors
Verified
Journal Article
Open access
PDF file
MLMing Lin
Ming Lin
JYJaewoo Yoon
Jaewoo Yoon
BKByeong-Woo Kim
Byeong-Woo Kim
Localization is one of the key components in the operation of self-driving cars. Owing to the noisy global positioning system (GPS) signal and multipath routing in urban environments, a novel, practical approach is needed. In this study, a sensor fusion approach for self-driving cars was developed. To localize the vehicle position, we propose a particle-aided unscented Kalman filter (PAUKF) algorithm. The unscented Kalman filter updates the vehicle state, which includes the vehicle motion model and non-Gaussian noise affection. The particle filter provides additional updated position measurement information based on an onboard sensor and a high definition (HD) map. The simulations showed that our method achieves better precision and comparable stability in localization performance compared to previous approaches.

14 April 2022 in Applied Sciences
Verified
Journal Article
Open access
PDF file
MPMinsick Park
Minsick Park
JLJong-Kyu Lee
Jong-Kyu Lee
BKByeong-Woo Kim
Byeong-Woo Kim
The state of health (SOH) of a lithium-ion battery (LIB) should be accurately estimated to ensure its safe operation under various driving conditions for electric vehicles. To estimate the SOH of LIBs, it is necessary to develop a technique that can properly process data with nonlinear characteristics related to the voltage and temperature of various electrochemical reactions in the batteries. To this end, we adopted wavelet transform methods to facilitate feature extraction for the preprocessing of nonlinear characteristic data from LIBs. Convolutional neural network (CNN) and long short-term memory (LSTM) techniques were used for the lithium-ion SOH estimation using the wavelet transform method. SOH estimation models using either the conventional data preprocessing technology normalization or wavelet transform method were compared. The SOH estimation accuracy of the model in which the wavelet transform was applied to the LSTM learning technology was 98.92%, and it was confirmed that the performance was improved compared with that of the normalization technology. In summary, this paper proposes a method that can improve the SOH estimation accuracy of LIBs compared to conventional methods by using data with nonlinear characteristics related to intrinsic changes in the voltage and temperature of LIBs.

23 June 2024 in Sensors
Verified
Journal Article
Open access
PDF file
JLJungEun Lee
JungEun Lee
BKByeong-Woo Kim
Byeong-Woo Kim
Given the increased significance of electric vehicles in recent years, this study aimed to develop a novel form of direct yaw-moment control (DYC) to enhance the driving stability of four-wheel independent drive (4WID) electric vehicles. Specifically, this study developed an innovative non-singular fast terminal sliding mode control (NFTSMC) method that integrates NFTSM and a fast-reaching control law. Moreover, this study employed a radial basis function neural network (RBFNN) to approximate both the entire system model and uncertain components, thereby reducing the computational load associated with a complex system model and augmenting the overall control performance. Using the aforementioned factors, the optimal additional yaw moment to ensure the lateral stability of a vehicle is determined. To generate the additional yaw moment, we introduce a real-time optimal torque distribution method based on the vertical load ratio. The stability of the proposed approach is comprehensively verified using the Lyapunov theory. Lastly, the validity of the proposed DYC system is confirmed by simulation tests involving step and sinusoidal inputs conducted using Matlab/Simulink and CarSim software. Compared to conventional sliding mode control (SMC) and NFTSMC methods, the proposed approach showed improvements in yaw rate tracking accuracy for all scenarios, along with a significant reduction in the chattering phenomenon in control torques.

1 January 2014 in Journal of Applied Mathematics
Verified
Journal Article
Open access
PDF file
HCHong Cho
Hong Cho
GKGyoung-Eun Kim
Gyoung-Eun Kim
BKByeong-Woo Kim
Byeong-Woo Kim
Conventional intelligent vehicles have performance limitations owing to the short road and obstacle detection range of the installed sensors. In this study, to overcome this limitation, we tested the usability of a new conceptual autonomous emergency braking (AEB) system that employs vehicle-to-vehicle (V2V) communication technology in the existing AEB system. To this end, a radar sensor and a driving and communication environment constituting the AEB system were simulated; the simulation was then linked by applying vehicle dynamics and control logic. The simulation results show that the collision avoidance relaxation rate of V2V communication-based AEB system was reduced compared with that of existing vehicle-mounted-sensor-based system. Thus, a method that can lower the collision risk of the existing AEB system, which uses only a sensor cluster installed on the vehicle, is realized.

4 February 2025 in Sensors
Verified
Journal Article
Open access
PDF file
JLJungEun Lee
JungEun Lee
BKByeong-Woo Kim
Byeong-Woo Kim
This study proposes an adaptive non-singular fast terminal sliding mode control (NFTSMC)-based direct yaw moment control (DYC) strategy to enhance driving stability in four-wheel independent drive (4WID) electric vehicles. Unlike conventional SMC, the proposed method dynamically adapts to system uncertainties and reduces chattering, a critical issue in control applications. The approach begins with the development of an NFTSMC method, analyzing its performance to identify areas for improvement. To enhance robustness and responsiveness, a novel adaptive NFTSMC method is introduced. This method integrates a non-singular fast terminal sliding mode surface with a novel adaptive fast-reaching control law that combines an adaptive switching mechanism and a fast-reaching law. The designed adaptive switching law adjusts the sliding gain in real time based on system conditions, reducing chattering without needing an upper bound on uncertainties as required by traditional NFTSMC methods. Concurrently, the fast-reaching law ensures rapid convergence from any initial condition and accurate tracking performance. Simulation results across various steering maneuvers, including step, sinusoidal, and fish-hook inputs, demonstrate that the proposed method significantly improves tracking accuracy and driving stability over traditional SMC and NFTSMC methods. Marked reductions in RMS and peak yaw rate errors, and effective chattering mitigation, highlight advancements in vehicle safety and stability.

16 October 2024 in IEEE Access
Verified
Journal Article
Open access
JAJinsu An
Jinsu An
DLDong Hee Lee
Dong Hee Lee
MPMuhamad Dwisnanto Putro
Muhamad Dwisnanto Putro
BKByeong-Woo Kim
Byeong-Woo Kim
Object detection using drones is a sophisticated technology that employs a camera mounted on a drone in conjunction with a computer vision algorithm to pinpoint the precise location of an object and ascertain its type. Drones are capable of rapidly scanning extensive areas, thereby facilitating efficient data collection and analysis. This capability can yield critical information and bolster swift response efforts. The utilization of object detection technology in drones offers numerous advantages. Nevertheless, despite the benefit of drones’ ability to swiftly scan wide areas, several challenges persist, including image resolution, the detection of small-sized objects, overlapping objects, and concentrated distributions. In this paper, we introduce DCE-YOLOv8, an advanced model based on YOLOv8. DCE-YOLOv8 is engineered to address the low detection rate of small objects in drone imagery. To effectively detect small objects, it is imperative to either enhance the resolution of drone images or efficiently extract the features of small objects. Additionally, the efficient integration of these extracted features is crucial. The Efficient Residual Bottleneck and Divided Context Extraction modules are incorporated into the Backbone, with the Efficient Residual Bottleneck reducing the number of parameters to render the model more lightweight. The Divided Context Extraction module focuses on extracting features pertinent to small objects. Subsequently, the rate of missed detections is mitigated by comprehensively merging the shallow and deep features extracted from the neck part. The proposed method is trained using the VisDrone Dataset and demonstrated superior detection rates compared to other cutting-edge methods. Compared to YOLOv8 small version, the mean Average Precision value improved by approximately 43%, increasing from 22.8mAP to 32.7mAP, while the number of parameters decreased by about 57%, from 11,166,560 to 4,822,382. The Average Inference Time per Image has been optimized to 11.4 ms, which, while slower than YOLOv8’s 5.9 ms, still maintains a robust frame rate of 87.71 FPS, underscoring its viability for real-time detection applications.