Jin‐Bae Park

ORCID: 0000-0003-3469-199X
Publications
Citations
Views
---
Saved
---
About
Contact & Profiles
Research Areas
  • Stability and Control of Uncertain Systems
  • Fuzzy Logic and Control Systems
  • Adaptive Control of Nonlinear Systems
  • Neural Networks and Applications
  • Advanced Algorithms and Applications
  • Target Tracking and Data Fusion in Sensor Networks
  • Inertial Sensor and Navigation
  • Cooperative Communication and Network Coding
  • Neural Networks Stability and Synchronization
  • Matrix Theory and Algorithms
  • Advanced MIMO Systems Optimization
  • Robotic Path Planning Algorithms
  • Electrical Fault Detection and Protection
  • Control Systems and Identification
  • Chaos control and synchronization
  • Infrared Target Detection Methodologies
  • Elasticity and Wave Propagation
  • Video Surveillance and Tracking Methods
  • Control and Dynamics of Mobile Robots
  • Advanced Sensor and Control Systems
  • Cybersecurity and Information Systems
  • Engineering and Test Systems
  • Robotic Locomotion and Control
  • Magnetic Bearings and Levitation Dynamics
  • Industrial Technology and Control Systems

Kyung Hee University
2012-2021

Hanwha Solutions (South Korea)
2020

Yonsei University
2008-2018

Pohang University of Science and Technology
1997

A typical semi-supervised learning-based scheme is based on training a single model for labeled data. For unlabeled data, it uses the pseudo-labeling method to obtain labels. However, samples during are often filtered using probability threshold, which suffers from challenge of effective threshold selection. In case high correct may not be labeled, and in low can wrongly labeled. This issue degrades overall performance model. paper addresses this vital by proposing novel approach SSL named...

10.1109/access.2021.3124200 article EN cc-by IEEE Access 2021-01-01

In prevalent knowledge distillation, logits in most image recognition models are computed by global average pooling, then used to learn encode the high-level and task-relevant knowledge. this work, we solve limitation of logit transfer distillation context. We point out that it prevents informative spatial information, which provides localized as well rich relational information across contexts an input scene. To exploit propose a simple yet effective approach. add local pooling layer branch...

10.1109/iccv48922.2021.00623 article EN 2021 IEEE/CVF International Conference on Computer Vision (ICCV) 2021-10-01

Visual target tracking is one of the major fields in computer vision system. Object has many practical applications such as automated surveillance system, military guidance, traffic management fault detection artificial intelligence and robot But it difficult to track objects with image sensor. Especially, multiple harder than single object tracking. This paper proposes algorithm based on Kalman filter. Our uses filter number moving frame. If exist image, however, we obtain measurements....

10.1109/sice.2014.6935235 article EN 2014-09-01

In this paper, a joint user association (UA) scheme with JP-CoMP using hybrid self-organizing network (SON) is proposed for practical clustered heterogeneous cellular (cHCN) to maximize the network-wide proportional fairness among users. The cell range expansion and enhanced intercell interference coordination have been considered as key items in long-term evolution-advanced offload macrocell users small-cell base stations (sBSs). However, cHCN where sBSs are not distributed at random but...

10.1109/tvt.2017.2748570 article EN IEEE Transactions on Vehicular Technology 2017-09-04

A practical pinch torque estimator based on the H <sub xmlns:mml="http://www.w3.org/1998/Math/MathML" xmlns:xlink="http://www.w3.org/1999/xlink">infin</sub> filter is proposed for low-cost anti-pinch window control systems. To obtain acceptable angular velocity measurements from Hall-effect sensor outputs, calculation algorithm with measurement noise reduction logics. Apart previous works or estimates detecting pinched condition, detection makes use of rate information. do this, augmented to...

10.1109/icit.2005.1600781 article EN 2006-03-10

Abstract In order to understand the relationship between processing conditions and properties of weld‐lines on a molecular level, it is necessary evaluate true strength weld‐line that not affected by V‐shape notch surface zone. this experiment, several brittle, ductile, or phase‐separated polymers was evaluated using compact tension test measuring critical stress intensity factor, K IC , J ‐value, results were compared with those obtained tensile testing. For brittle such as poly(methyl...

10.1002/pen.11766 article EN Polymer Engineering and Science 1997-07-01

A practical pinch torque estimator based on the Kalman filter is proposed for low-cost anti-pinch window control systems. To obtain accurate angular velocity from Hall-effect sensor measurements, calculation algorithm by including measurement noise reduction logics. Apart previous works or estimates detecting pinched condition, detection makes use of rate information. do this, augmented to system model and derived applying steady-state recursion model. The motivation this approach comes idea...

10.1109/iecon.2005.1568915 article EN 2005-01-01

A one‐dimensional (1D) virtual force field (VFF) algorithm for real‐time reflexive local path planning of mobile robots is proposed. The 1D‐VFF composed the steering, obstacle and integrated fields (IFFs). steering (SFF) generated by or global goal position. This SFF leads a robot to goal. (OFF) created raw data range measurement sensor (RMS). By this OFF, avoids obstacles. IFF produced combining OFFs in which weights between 0 1 are multiplied. Through IFF, final command reaches avoiding...

10.1049/el.2014.2787 article EN Electronics Letters 2014-09-01

본 논문에서는 선도 로봇을 추종하는 군집 로봇의 대형 제어를 인공 포텐셜 장을 사용하여 제안한다. 또한, 장은 물리적으로 해석하기 쉬운 전기장을 모델링하여 구성하고, 장애물을 더욱 효과적으로 모델링하기 위해서, 장애물의 모양에 따라 전기장의식을 달리한다. 제안하는 방법은 경로를 통해 계획한 뒤, 추종 로봇이 뒤따라가는 형태로 구성된다. 마지막으로 시뮬레이션 예제를 기법의 타당성을 검증한다. In this paper, artificial potential field(APF) is applied to formation control for the leader-following swarm robot. Furthermore, APF constructed by applying electrical field model. Moreover, model obstacle effectively, each has different form due equation. The proposed...

10.5391/jkiis.2012.22.4.476 article EN Journal of Korean institute of intelligent systems 2012-08-25

Visual target tracking is one of the important issues in computer vision. Especially, particular object an environment multiple objects has many practical applications such as automated surveillance system and robot vision system. However, it difficult to track a because problems are occurred merging splitting. This paper proposes algorithm when similar present. based on Kalman filter. But, standard filter requires measurement for state estimation. if exist image, we obtain measurements...

10.1109/iccas.2013.6703957 article EN 2013-10-01

The global energy demand requires alternative resource from the fossil fuels and nuclear fission. Ocean should take in charge of a key role renewable due to high density vast amount its near future. In particular, an oscillating water column-wave converter (OWC-WEC) is one most promising devices capture ocean waves. This kind consists chamber, power off (PTO) module such as Wells turbine or impulse turbine, generator. An overall efficiency this device determined by each component: wave...

10.1109/iccas.2015.7364693 article EN 2015-10-01

본 논문에서는 4족 로봇의 허리 관절을 이용하여 효율적인 경사면 보행을 위한 걸음새 생성 방법을 제안한다. 갖는 기구학적 모델은 Denavit-Hartenberg 표현 방법과 대수적 유도하고, 다리 이동 순서는 물결 걸음새(wave gait)를 사용한다. 한편 제안한 방법에서는 기구적 제한과 보폭의 감소를 완화하기 위해 경사면의 경사도에 따라 적절한 상체 및 하체의 관절각을 결정하고, 에너지 안정도 여유(energy stability margin)를 증가시키기 도달 영역(workspace)의 탐색을 통해 발끝 위치를 결정한다. 마지막으로, 컴퓨터 모의 실험을 논문에서 알고리듬의 효용성 실제 적용 가능성을 검증한다. In this paper, we propose a gait generation method for quadruped robot to walk efficiently on the slope, which uses waist joint of robot. We derive...

10.5391/jkiis.2012.22.5.617 article EN Journal of Korean institute of intelligent systems 2012-10-25

This paper proposes a new linear recursive target state and time-to-collision estimator for the development of automotive collision warning system. The addressed problem can be cast into representative nonlinear estimation under cluttered environment. To prevent tracking performance degradation due to inherent nonlinearity between polar coordinates measurements radar state, practical filter design scheme employing estimated line-of-sight (LOS) Cartesian coordinate system (ELCCS) is proposed....

10.1109/iecon.2011.6119381 article EN IECON 2020 The 46th Annual Conference of the IEEE Industrial Electronics Society 2011-11-01

본 논문에서는 하이브리드 발전 시스템에서의 최대 전력 추종을 위한 디지털 퍼지 제어기 설계를 목표로 한다. 시스템은 풍력과 태양광, 두 개의 시스템으로 구성된다. 각 발전기에서 전압과 전류는 일반적으로 비선형 관계에 있기 때문에, 모델 기반 제어기를 사용하여 비선형성을 효율적으로 제어하게 된다. 또한 마이크로프로세서 기반제어 시스템의 구축을 위하여 최신 재설계 기법을 사용하며, 설계하게 마지막으로 플랜트를 통한 실험으로써 제안 기법의 우수성을 입증한다. We present the development of digital fuzzy controller for maximum power regulation. A hybrid system that comprises wind and photovoltaic generation subsystems, battery bank is developed in this paper. use Takaki-Sugeno (T-S) model to deal...

10.5391/jkiis.2006.16.6.753 article EN Journal of Korean institute of intelligent systems 2006-12-25

본 논문에서는 퍼지 이론을 이용한 풍력 터빈의 변화 속도 제어에 관해 다루고자 한다. 일반적인 속도는 복잡한 비선형성으로 나타내어지며, 플랜트를 구성하는 각 파라미터의 수치 역시 불확실하다. 이와 같은 복잡성을 해결하기 위하여, 우리는 비선형성 및 불확실성에 강인한 제어 제안하고자 우선 속도에 대한 정확한 모델링을 수행하게 된다. 그리고 재해석된 Takagi-Sugeno (T-S) 모델에 적합한 제어기를 설계하게 되며, 리아푸노프 안정도에 기반한 시스템의 안정도를 증명하게 마지막으로, 가상 시뮬레이션을 통해 제안된 기법의 효율성을 입증하게 In this paper, we present the robust fuzzy algorithm for variable speed control of wind turbines. Generally, plants turbines are consisted complex nonlinearities, and parameters represented...

10.5391/jkiis.2008.18.1.001 article EN Journal of Korean institute of intelligent systems 2008-02-25

본 논문은 가우시안 정보원을 전송할 때 평균 왜곡이 줄어들도록 두 레이어 (layer) 중첩 부호와 순차 복호를 적용한 단일 릴레이 노드 협력 통신 시스템을 고려한다. 이 위해 제안하는 전송 방식은 노드가 자신의 복호결과와 피드백 정보인 목적 노드의 복호결과에 따라 알맞은 신호를 구성하여 전송하는 것이다. 신호는 복호-후 전달과 증폭-후-전달 기법을 혼합하여 설계함으로써 최종 복호 시의 아웃티지 확률을 줄이도록 한다. 제안 기법의 성능은 확률 분석 결과를 이용하여 다양한 위치에서의 최적 왜곡 성능으로 살펴본다. 그 결과 대부분의 위치에서 방식이 기존 방식보다 성능이 좋으며, 특히 원천 노드에 가까울수록 성능 이득이 더욱 증가함을 볼 수 있다. This paper considers a cooperative communication system with single relay node, where two-layer superposition coding and successive...

10.7840/kics.2012.37.7b.517 article EN The Journal of Korean Institute of Communications and Information Sciences 2012-07-31

본 논문에서는 접근 각도 개념을 이용하여 과소작동기 형태의 무인 수중 잠수정의 경로 추적 제어기 설계 방법을 제안한 다. 과소 작동기 형태를 가지는 잠수정은 종 방향 추진력과 회전력에 의해 속도와 방향을 조절하나, 횡 추 진기가 없기 때문에 방향에 대한 움직임을 제어 할 수 없다. 이러한 문제를 해결하기 위하여 기준 경로에 제안하고, 각도를 제어기를 설 계한다. 이를 위해 동체 고정 좌표계에서 새로운 오차 방정식을 구하고, 리아푸노프 기반으로 설계한다. 컴퓨터 시뮬레이션 통해 방법에 설계된 제어기의 성능을 검증한다. In this paper, we propose a method for designing the path tracking controller using an approach angle concept underactuated autonomous underwater vehicle (AUV). The AUV is controlled by surge speed and...

10.5391/jkiis.2012.22.2.225 article EN Journal of Korean institute of intelligent systems 2012-04-25

본 논문에서는 무선 센서 네트워크를 위한 비선형 네트워크 제어 시스템의 출력 궤환 분산 퍼지 제어기를 설계한다. 특히나, 제어를 시스템은 출력의 패킷 손실과 입력 전송 실패를 가진다고 가정한다. 제어기 설계를 위해 먼저 하위 Takagi-Sugeno (T-S) 모델을 제시하고, 각 시스템에 대한 제안된 포함한 폐루프 안정도 조건을 Lyapunov 방정식을 통하여 구하고, 구해진 선형 행렬 부등식으로 나타내어, 이를 통해 제어기의 이득값을 구한다. 모의실험을 효용성을 평가한다. In this paper, a decentralized fuzzy output feedback controller for the nonlinear networked control system is proposed wireless sensor network. Especially, it assumed that has packet loss and input transmission failure. For of...

10.5391/jkiis.2009.19.3.323 article EN Journal of Korean institute of intelligent systems 2009-06-25

This paper presents intelligent digital redesign (IDR) method of global approach for fuzzy-model-based controllers. For effectiveness and stabilization continuous-time uncertain nonlinear systems under discrete-time controller, Takagi-Sugeno (T-S) fuzzy model is used to represent the complex system. And design problems viewed as a convex optimization problem that we minimize error norm bounds between nonlinearly interpolated linear operators be matched. Also, by using power series, analyzed...

10.1109/sice.2006.314822 article EN 2006 SICE-ICASE International Joint Conference 2006-01-01

In this paper, the Bayesian recurrent neural network is proposed to predict time series data. A predictor requests proper learning strategy adjust weights, and one needs prepare for non-linear non-stationary evolution of weights. The in paper estimates not single set weights but probability distributions other words, vector as a state space method, its are estimated accordance with particle filtering process. This approach makes it possible obtain more exact estimation aspect architecture,...

10.5302/j.icros.2004.10.12.1295 article EN Journal of Control Automation and Systems Engineering 2004-12-01

This paper proposes an identification method using a self recurrent wavelet neural network (SRWNN) for dynamic systems. The architecture of the proposed SRWNN is modified model (WNN). But, unlike WNN, since mother layer composed self-feedback neurons, has ability to store past information wavelet. Thus, in architecture, used identifying nonlinear gradient descent with adaptive teaming rates (ALRs) applied 1.am parameters identifier (SRWNNI). ALRs are derived from discrete Lyapunov stability...

10.5302/j.icros.2005.11.9.781 article EN Journal of Control Automation and Systems Engineering 2005-09-01
Coming Soon ...