• 제어로봇시스템학회:학술대회논문집
• /
• 제어로봇시스템학회 1995년도 Proceedings of the Korea Automation Control Conference, 10th (KACC); Seoul, Korea; 23-25 Oct. 1995
• /
• pp.47-50
• /
• 1995

This paper concerns an adaptive control scheme which is an extension of the simplified adaptive control. Originally, the SAC approach was developed based on the command generator tracker (CGT) theory for perfect model tracking. An attractive point of the SAC is that a control input can be synthesized without any prior knowledge about plant structure. However, a feedforward dynamic compensator of the CGT is removed from the basic structure of the SAC. This deletion of the compensator makes perfect model tracking difficult against even a step input. In this paper, an adaptive control system is redesigned to achieve perfect model tracking for as long as possible by reviving the dynamic compensator of the CGT. The proposed method is applied to slewing control of a flexible space structure and compared to the SAC responses.

• 한국CDE학회논문집
• /
• 제22권1호
• /
• pp.28-36
• /
• 2017

Sharing of CAD data plays a key role in the PLM and a lightweight model is widely used for visualizing and sharing a large data. The lightweight model is mainly composed of triangular elements to minimize file size. There is no problem at all to visually confirm the shape based on these triangular elements but there is a limit to numerically calculate the exact position on the curve or surface. In this paper, a boundary curve generation method using triangular elements is proposed to increase the utilization of lightweight models. After matching connectivity of triangular elements, boundary element edges are extracted. Boundary curves are generated by connecting of these boundary element edges. This proposed method has been tested on several models to demonstrate the feasibility.

• 한국정밀공학회지
• /
• 제21권1호
• /
• pp.124-132
• /
• 2004

A hydraulic system is difficult to keep the performance due to non-linearity, load pressure which changes according to working condition and system parameter variation, the requirement of control algorithm has been risen in order to satisfy them. An adaptive control is a control method which is suggested to achieve a control object though plant characteristics change. In spite of the case that plant characteristics and the degree of variation are difficult to grasp, adaptive control can keep the characteristics of closed-loop system regularly. In this study GMVAC(generalized minimum variance adaptive control) combined with output error feedback is proposed in order to solve problems of non-minimum phase, vibration and overshoot in initial response of the plant. The control performance according to the variation of characteristics of the plant is evaluated by changing the supply pressure only.

• 전자공학회논문지SC
• /
• 제47권4호
• /
• pp.7-14
• /
• 2010

유한 상태 머신으로 표현되는 비동기 머신의 안정 상태 동작은 피드백 제어를 통해서 원하는 목적에 맞게 교정될 수 있다. 본 논문에서는 불확실한 상태 천이를 가지는 입력/상태 비동기 머신을 위한 상태 피드백 제어기를 제안한다. 비동기 머신은 결정적인 동작 특성을 가지고 있으나 모델 불확실성, 내부 고장 등으로 인해서 일부 영역의 상태 천이 함수가 불확실하다. 교정 제어의 목적은 불확실한 상태 천이를 고려하면서 머신의 폐루프 동작이 주어진 정상적인 모델 동작의 부분 집합이 되도록하는 일이다. 또 제어기는 실제 교정 동작을 수행하면서 획득하는 머신의 정확한 상태 천이를 그 다음 제어 동작에 반영한다. 즉 학습을 통해서 폐루프 시스템이 이룰 수 있는 모델의 부분 동작의 범위가 더 확대된다. 사례 연구를 제안된 제어기의 설계 과정을 예시한다.

• 응용통계연구
• /
• 제36권5호
• /
• pp.415-445
• /
• 2023

생존 자료에서 Hade 등 (2020) 은 시간-의존 교란 변수가 환자의 처치 시점에 영향을 미칠 때, 해당 효과를 보정하여 treatment delay effect를 올바르게 추정하기 위해 성향 점수 매칭 방법을 이용하였다. 이 때, treatment delay effect란 환자가 관심 있는 지연 시점만큼 늦게 처치를 받는 경우 제 때 받는 경우에 비해 사건 발생 위험에 미치는 영향을 의미한다. 본 연구에서는 또 다른 성향 점수 기반 모형인 Cox-MSM 모형 또한 해당 효과를 올바르게 추정할 수 있는지 모의 실험을 통해 확인 및 기존 매칭 모형과 비교하였다. 모의실험 결과, 세 가지 모형 모두 다양한 시나리오 내에서 treatment delay effect를 올바르게 추정함을 확인하였다. 특히 모든 시나리오 내에서 Cox-MSM의 제곱근평균제곱오차의 값이 가장 낮았으며, restricted Cox matching 모형에서 가장 큰 값을 가지는 것으로 나타났다. 결론적으로, 성향 점수에 기반하나 매칭이 아닌 방법 또한 treatment delay effect 적용이 가능하다는 결과를 제공한다. 추후 G-formula과 같이 성향 점수 기반이 아닌 모형에서도 적용이 가능한지에 대한 상세 연구가 필요하다고 사료된다.

• ETRI Journal
• /
• 제41권6호
• /
• pp.771-781
• /
• 2019

We propose the VS³-NET model to solve the task of question answering questions with machine-reading comprehension that searches for an appropriate answer in a given context. VS³-NET is a model that trains latent variables for each question using variational inferences based on a model of a simple recurrent unit-based sentences and self-matching networks. The types of questions vary, and the answers depend on the type of question. To perform efficient inference and learning, we introduce neural question-type models to approximate the prior and posterior distributions of the latent variables, and we use these approximated distributions to optimize a reparameterized variational lower bound. The context given in machine-reading comprehension usually comprises several sentences, leading to performance degradation caused by context length. Therefore, we model a hierarchical structure using sentence encoding, in which as the context becomes longer, the performance degrades. Experimental results show that the proposed VS³-NET model has an exact-match score of 76.8% and an F1 score of 84.5% on the SQuAD test set.

• JSTS:Journal of Semiconductor Technology and Science
• /
• 제2권1호
• /
• pp.59-69
• /
• 2002

A semi-analytical model based on exact numerical analysis of the 2DEG channel in pseudo-morphic HEMT (PMHEMT) is presented. The exactness of the model stems from solving both Schrodinger's wave equation and Poisson's equation simultaneously and self-consistently. The analytical modeling of the device terminal characteristics in relation to the charge control model has allowed a best fit with the geometrical and structural parameters of the device. The numerically obtained data for the charge control of the channel are best fitted to analytical expressions which render the problem analytical. The obtained good agreement between experimental and modeled current/voltage characteristics and small signal parameters has confirmed the validity of the model over a wide range of biasing voltages. The model has been used to compare both the performance and characteristics of a PMHEMT with a competetive HEMT. The comparison between the two devices has been made in terms of 2DEG density, transfer characteristics, transconductance, gate capacitance and unity current gain cut-off frequency. The results show that PMHEMT outperforms the conventional HEMT in all considered parameters.

• 제어로봇시스템학회논문지
• /
• 제12권4호
• /
• pp.389-394
• /
• 2006

Inaccurate localization exposes a robot to many dangerous conditions. It could make a robot be moved to wrong direction or damaged by collision with surrounding obstacles. There are numerous approaches to self-localization, and there are different modalities as well (vision, laser range finders, ultrasonic sonars). Since sensor information is generally uncertain and contains noise, there are many researches to reduce the noise. But, the correctness is limited because most researches are based on statistical approach. The goal of our research is to measure more exact robot location by matching between built VRML 3D model and real vision image. To determine the position of mobile robot, landmark-localization technique has been applied. Landmarks are any detectable structure in the physical environment. Some use vertical lines, others use specially designed markers, In this paper, specially designed markers are used as landmarks. Given known focal length and a single image of three landmarks it is possible to compute the angular separation between the lines of sight of the landmarks. The image-processing and neural network pattern matching techniques are employed to recognize landmarks placed in a robot working environment. After self-localization, the 2D scene of the vision is overlaid with the VRML scene.

• 대한기계학회논문집A
• /
• 제23권11호
• /
• pp.1929-1939
• /
• 1999

This paper presents a neural network approach, which was named PRONET, to 3D object recognition and pose calculation. 3D objects are represented using a set of centroidal profile patterns that describe the boundary of the 2D views taken from evenly distributed view points. PRONET consists of the training stage and the execution stage. In the training stage, a three-layer feed-forward neural network is trained with the centroidal profile patterns using an error back-propagation method. In the execution stage, by matching a centroidal profile pattern of the given image with the best fitting centroidal profile pattern using the neural network, the identity and approximate orientation of the real object, such as a workpiece in arbitrary pose, are obtained. In the matching procedure, line-to-line correspondence between image features and 3D CAD features are also obtained. An iterative model posing method then calculates the more exact pose of the object based on initial orientation and correspondence.

• 대한전기학회:학술대회논문집
• /
• 대한전기학회 2006년도 심포지엄 논문집 정보 및 제어부문
• /
• pp.318-320
• /
• 2006

Inaccurate localization exposes a robot to many dangerous conditions. It could make a robot be moved to wrong direction or damaged by collision with surrounding obstacles. There are numerous approaches to self-localization, and there are different modalities as well (vision, laser range finders, ultrasonic sonars). Since sensor information is generally uncertain and contains noise, there are many researches to reduce the noise. But, the correctness is limited because most researches are based on statistical approach. The goal of our research is to measure more exact robot location by matching between built VRML 3D model and real vision image. To determine the position of mobile robot, landmark-localitzation technique has been applied. Landmarks are any detectable structure in the physical environment. Some use vertical lines, others use specially designed markers, In this paper, specially designed markers are used as landmarks. Given known focal length and a single image of three landmarks it is possible to compute the angular separation between the lines of sight of the landmarks. The image-processing and neural network pattern matching techniques are employed to recognize landmarks placed in a robot working environment. After self-localization, the 2D scene of the vision is overlaid with the VRML scene.