• Journal of Electrical Engineering and Technology
• /
• v.10 no.3
• /
• pp.925-934
• /
• 2015

This paper considers the robust controller design problem for a boost DC-DC converter. Based on the Takagi-Sugeno fuzzy model-based approach, a fuzzy controller as well as a fuzzy load conductance observer are designed. Sufficient conditions for the existence of the controller and the observer are derived using Linear Matrix Inequalities (LMIs). LMI parameterizations of the gain matrices are obtained. Additionally, LMI conditions for the existence of the fuzzy controller and the fuzzy load observer guaranteeing α-stability, quadratic performance are derived. The exponential stability of the augmented fuzzy observer-controller system is shown. It is also shown that the fuzzy load observer and the fuzzy controller can be designed independently. Finally, the effectiveness of the proposed method is verified via experimental and simulation results under various conditions.

• 제어로봇시스템학회:학술대회논문집
• /
• 2000.10a
• /
• pp.25-25
• /
• 2000

This paper presents a design method of sliding mode control (SMC) for single input linear systems with mismatched uncertainties. We define a virtual state based on the controllable canonical form of the nominal system. And we define a sliding surface for the augmented system with a virtual state. This sliding surface makes it possible to use SMC technique with various types of controllers. In this paper, we construct a controller that combines SMC with robust controller. We design a robust controller for the system with only mismatched uncertainties using a form of linear matrix inequality (LMI). We make a virtual state from this robust control input and the states of the nominal system. And we design a sliding mode controller that stabilizes the overall closed-loop system.

• Journal of the Korean Society for Precision Engineering
• /
• v.15 no.6
• /
• pp.145-152
• /
• 1998

This paper presents an optimum design of a rotor-bearing spindle system for a ultra centrifuge (80,000 RPM) supported by ball bearings with nonlinear stiffness characteristics. To obtain the nonlinear bearing stiffnesses, a ball bearing is modeled in five degrees of freedom and is analyzed quasi-statically. The dynamic behaviors of the nonlinear rotor-bearing system are analyzed by using a transfer-matrix method iteratively. For optimization. we use the cost function that simultaneously minimizes the weight of a rotor and maximizes the separation margins to yield the critical speeds as far from the operating speed as possible. Augmented Lagrange Multiplier (ALM) method is employed for the nonlinear optimization problem. The result shows that the rotor-bearing spindle system is optimized to obtain 9.5％ weight reduction and 21％ separation margin.

• Journal of Institute of Control, Robotics and Systems
• /
• v.7 no.3
• /
• pp.193-197
• /
• 2001

This paper presents a design method of sliding model control (SMC) for single input linear systems with mismatched uncertainties. We define a virtual state based on the controllable canonical form of the nominal system. And we defined a sliding surface for the augmented system with a virtual state. This sliding surface makes it possible to use the SMC technique with various types of controllers. In this paper, we construct a controller that combines SMC with robust controller. We design a robust controller for the system with mismatched uncertainties using a form of linear matrix inequality(LMI). We make a virtual state from this robust control input and the states of the nominal system. And we design a sliding model controller that stabilizes the overall closed-loop system.

• Proceedings of the Korean Institute of Navigation and Port Research Conference
• /
• 2011.06a
• /
• pp.161-163
• /
• 2011

최근 몇 년사이 CCTV에 대한 활용도와 중요성이 대두되고 있으며, CCTV의 해상도 또한 SD급 CCTV에서 HD급 CCTV로 대부분 전환되고 있는 상황이다. HD급 CCTV는 고화질, 고해상도를 제공하는 장점을 갖지만, 많은 데이터 량으로 인해 실시간 처리가 어려운 문제점 또한 포함하고 있다. 또한 CCTV는 지능형 CCTV로의 기술적인 진보를 거듭하고 있으며, 대표적인 solution으로 증강현실(AR:Augmented Reality)을 꼽을 수 있다. 본 논문에서는 HD-CCTV의 최대 장점이라고 할 수 있는 실시간 영상에 선박의 정보(AIS : Automatic Identification System)을 결합하여 실시간으로 항만을 모니터링 하는 시스템을 구현하였다. 또한 750mm Zoom Lens를 탑재한 PanTilt 장비를 이용하여 선박을 targeting 하는 시스템 또한 구현하였다. 증강현실을 이용하여 실시간 영상과 선박 정보를 결합하였으며, 이를 구현하기 위해 perspective projection 방법을 통해 3차원 공간좌표계를 2차원으로 투영하였다. 실시간 처리를 위해 입력영상을 Block으로 분할하여 목적 좌표를 검색하였으며, 선박정보의 부드러운 이동을 위해 Dead Reckon 기법과 linear prediction 방법을 이용하여 선박 위치를 예측하였다. 마지막으로 삼각측량법에 기반하여 현재 PanTilt 장비를 목적된 장소로 이동시키는 시스템을 구현하였다.

• The Transactions of The Korean Institute of Electrical Engineers
• /
• v.60 no.11
• /
• pp.2119-2130
• /
• 2011

This paper proposes new delay-dependent robust stability criteria for neural networks with time-varying delays. By construction of a suitable Lyapunov-Krasovskii's (L-K) functional and use of Finsler's lemma, new stability criteria for the networks are established in terms of linear matrix inequalities (LMIs) which can be easily solved by various effective optimization algorithms. Two numerical examples are given to illustrate the effectiveness of the proposed methods.

• KSII Transactions on Internet and Information Systems (TIIS)
• /
• v.16 no.7
• /
• pp.2407-2424
• /
• 2022

Recently, Transformer has made great progress in video retrieval tasks due to its high representation capability. For the structure of a Transformer, the cascaded self-attention modules are capable of capturing long-distance feature dependencies. However, the local feature details are likely to have deteriorated. In addition, increasing the depth of the structure is likely to produce learning bias in the learned features. In this paper, an improved Transformer structure named TransDCS (Transformer with Dynamic Convolution and Shortcut) is proposed. A Multi-head Conv-Self-Attention module is introduced to model the local dependencies and improve the efficiency of local features extraction. Meanwhile, the augmented shortcuts module based on a dual identity matrix is applied to enhance the conduction of input features, and mitigate the learning bias. The proposed model is tested on MSRVTT, LSMDC and Activity-Net benchmarks, and it surpasses all previous solutions for the video-text retrieval task. For example, on the LSMDC benchmark, a gain of about 2.3% MdR and 6.1% MnR is obtained over recently proposed multimodal-based methods.

• Journal of Korea Multimedia Society
• /
• v.9 no.4
• /
• pp.421-433
• /
• 2006

This paper proposes a method to insert virtual objects into a real video stream based on feature tracking and camera pose estimation from a set of single-camera video frames. To insert or modify 3D shapes to target video frames, the transformation from the 3D objects to the projection of the objects onto the video frames should be revealed. It is shown that, without a camera calibration process, the 3D reconstruction is possible using multiple images from a single camera under the fixed internal camera parameters. The proposed approach is based on the simplification of the camera matrix of intrinsic parameters and the use of projective geometry. The method is particularly useful for augmented reality applications to insert or modify models to a real video stream. The proposed method is based on a linear parameter estimation approach for the auto-calibration step and it enhances the stability and reduces the execution time. Several experimental results are presented on real-world video streams, demonstrating the usefulness of our method for the augmented reality applications.

• The KIPS Transactions:PartB
• /
• v.13B no.5 s.108
• /
• pp.499-506
• /
• 2006

This Paper describes a fast and stable camera pose estimation method for real-time augmented reality systems. From the feature tracking results of a marker on a single frame, we estimate the camera rotation matrix and the translation vector. For the camera pose estimation, we use the shape factorization method based on the scaled orthographic Projection model. In the scaled orthographic factorization method, all feature points of an object are assumed roughly at the same distance from the camera, which means the selected reference point and the object shape affect the accuracy of the estimation. This paper proposes a flexible and stable selection method for the reference point. Based on the proposed method, we implemented a video augmentation system that inserts virtual 3D objects into the input video frames. Experimental results showed that the proposed camera pose estimation method is fast and robust relative to the previous methods and it is applicable to various augmented reality applications.

• Journal of Service Research and Studies
• /
• v.12 no.4
• /
• pp.50-71
• /
• 2022

After the declaration of the Covid-19 pandemic impacted most of the industries resulting economic fallout. Firms sought for solutions of governments regulations to prevent spread of infectious diseases. This led to demand rise of digital layer and spectrums of virtual reality. Replacing the reality in to virtual and interactions with the digital contents by augmented reality, the consequences were decrement of human-to-human contact. Concerns of Covid-19 and public interests of digital solutions has led to significant amounts of research and developments of Virtual/Augment Reality resulted to driven up new terms of extended reality. However, the uses in industries and the characteristics of the extended reality are currently not defined. In this paper the goal is to define and classify the uses and characteristics of extended reality based on previous researches suggested by research institute. By developing a new classification models of extended realities core technology, uses of industries and utility to analyze trends of extended reality. Two separate classification models of uses of industries and utility will be used as a tool by creating a linkage matrix. The x-axis is divided by utiliy classification model of extended reality. The y-axis are divided into classification model of uses in industries. This matrix will be used as a tool to present a guideline for industry-utility development where extended reality can be served as a service