• 제어로봇시스템학회:학술대회논문집
• /
• 제어로봇시스템학회 2005년도 ICCAS
• /
• pp.994-998
• /
• 2005

A new control method, called a simple model matching, has been recently developed by the author. This is very simple and be applied for linear and nonlinear discrete time systems with/without time lag. Based on this formulation, identification is examined in this paper using an interconnected neural network with the EBP-EWLS learning algorithm. With this result, a control method is also presented for a nonlinear discrete time system.

• ETRI Journal
• /
• 제41권3호
• /
• pp.371-382
• /
• 2019

Machine reading comprehension is the task of understanding a given context and finding the correct response in that context. A simple recurrent unit (SRU) is a model that solves the vanishing gradient problem in a recurrent neural network (RNN) using a neural gate, such as a gated recurrent unit (GRU) and long short-term memory (LSTM); moreover, it removes the previous hidden state from the input gate to improve the speed compared to GRU and LSTM. A self-matching network, used in R-Net, can have a similar effect to coreference resolution because the self-matching network can obtain context information of a similar meaning by calculating the attention weight for its own RNN sequence. In this paper, we construct a dataset for Korean machine reading comprehension and propose an $S^2-Net$ model that adds a self-matching layer to an encoder RNN using multilayer SRU. The experimental results show that the proposed $S^2-Net$ model has performance of single 68.82% EM and 81.25% F1, and ensemble 70.81% EM, 82.48% F1 in the Korean machine reading comprehension test dataset, and has single 71.30% EM and 80.37% F1 and ensemble 73.29% EM and 81.54% F1 performance in the SQuAD dev dataset.

• International Journal of Automotive Technology
• /
• 제5권2호
• /
• pp.135-144
• /
• 2004

In this paper, the model-matching control in the longitudinal autonomous driving system is investigated by vehicle dynamics simulation, which contains nonlinear subcomponents and simplified subcomponents. The design of the robust model-matching controller is performed by the characteristics of the 2 degrees of freedom controller, which is composed of the feedforward compensator and the feedback compensator. It makes the characteristics of tractive and brake force to be equivalent to the specific transfer function, which is suggested as the reference model. Mathematical models of vehicle dynamic analysis including the model-matching control are constructed for computer simulation. Then, simple examples on open-loop simulation without any controller and closed loop simulation with the model-matching controller are applied to check the validity of the robust controller. As the practical example, the autonomous driving system in the longitudinal direction is adopted. It is proved that the model-matching control is effective and adequate to the disturbances and the perturbations, which are shown in the responses of the change of a vehicle mass and a road gradient.

• ETRI Journal
• /
• 제34권2호
• /
• pp.292-295
• /
• 2012

Beyond the simple keyword matching methods in contextual advertising, we propose a rich contextual matching (CM) model adopting a classification method for topic targeting and a query expansion method for semantic ad matching. This letter reports on an investigation into the empirical effects of the CM model by comparing the click-through rates (CTRs) of two practical online news advertising systems. Based on the evaluation results from over 100 million impressions, we prove that the average CTR of our proposed model outperforms that of a traditional model.

• Asia Marketing Journal
• /
• 제17권4호
• /
• pp.89-104
• /
• 2016

Based on the observation that insurance companies in Korea, unlike those in other financial sectors and those in other countries, dominantly use the agent-based push-type marketing strategy, this paper hypothesizes that difference in distribution systems originating from characteristics of financial products can lead to welfare asymmetry between financial institutions and customers, merely due to their financial matching. For this analysis, we employ a simple matching theoretic model, try to understand the welfare implications of distribution systems from a matching theoretic perspective, and analyze the bottom of negative perceptions of insurance industry. The proposed model suggests that this welfare asymmetry derives mainly from financial matching through the distribution systems, which implies that any efforts to improve the insurance industry must consider changes in the matching process, namely the distribution system. We hope that this paper complements and extends the existing literature on insurance distribution systems in terms of methodologies and research subjects.

• 소음진동
• /
• 제9권2호
• /
• pp.340-347
• /
• 1999

The acoustic performance of reactive type single expansion chamber can be calculated theoretically by plane wave theory. But higher order model should be considered to widen the frequency range. Mode matching method has been developed to consider higher order modes, but very complicated algebra should be used. Munjal suggested a numerical collocation method, which can overcome the shortcomings of mode matching method, using the compatibility conditions for acoustic pressure and particle velocity at the junctions of area discontinuities. But the restriction of Munjal's method is that the ratio between the area of inlet(or outlet) pipe and that of chamber must be natural number. In this paper, the new method was suggested to overcome the shortcomings of Munjal's method. The predictions by this method was also compared with those by the finite element method and Munjal's method in order to demonstrate the accuracy of the modified method presented here.

• 제어로봇시스템학회:학술대회논문집
• /
• 제어로봇시스템학회 2005년도 ICCAS
• /
• pp.979-984
• /
• 2005

In this study, we suggest a tilt rotor aircraft and attempt to apply a nonlinear model matching control method for its maneuver. The proposed method is very simple and useful to construct the control law for the complicated nonlinear system such as aircraft motion.

• 대한전기학회:학술대회논문집
• /
• 대한전기학회 2007년도 심포지엄 논문집 정보 및 제어부문
• /
• pp.304-306
• /
• 2007

Methods of Object recognition from camera image are to compare features of color. edge or pattern with model in a general way. SIFT(scale-invariant feature transform) has good performance but that has high complexity of computation. Using simple color histogram has low complexity. but low performance. In this paper we represent a model as a color cooccurrence histogram. and we improve performance using pyramid matching. The color cooccurrence histogram keeps track of the number of pairs of certain colored pixels that occur at certain separation distances in image space. The color cooccurrence histogram adds geometric information to the normal color histogram. We suggest object recognition by pyramid matching of color cooccurrence histogram.

• 대한전기학회논문지:시스템및제어부문D
• /
• 제53권4호
• /
• pp.277-285
• /
• 2004

We present a new hybrid stereo matching technique in terms of the co-operation of area-based stereo and feature-based stereo. The core of our technique is that feature matching is carried out by the reference of the disparity evaluated by area-based stereo. Since the reference of the disparity can significantly reduce the number of feature matching combinations, feature matching error can be drastically minimized. One requirement of the disparity to be referenced is that it should be reliable to be used in feature matching. To measure the reliability of the disparity, in this paper, we employ the self-consistency of the disunity Our suggested technique is applied to the detection of 3D line segments by 2D line matching using our hybrid stereo matching, which can be efficiently utilized in the generation of the rooftop model from urban imagery. We carry out the experiments on our hybrid stereo matching scheme. We generate synthetic images by photo-realistic simulation on Avenches data set of Ascona aerial images. Experimental results indicate that the extracted 3D line segments have an average error of 0.5m and verify our proposed scheme. In order to apply our method to the generation of 3D model in urban imagery, we carry out Preliminary experiments for rooftop generation. Since occlusions are occurred around the outlines of buildings, we experimentally suggested multi-image hybrid stereo system, based on the fusion of 3D line segments. In terms of the simple domain-specific 3D grouping scheme, we notice that an accurate 3D rooftop model can be generated. In this context, we expect that an extended 3D grouping scheme using our hybrid technique can be efficiently applied to the construction of 3D models with more general types of building rooftops.

• 공학논문집
• /
• 제2권1호
• /
• pp.63-71
• /
• 1997

본 논문에서는 range 데이터로부터 평면, 또는 곡면을 갖는 물체의 표면을 인식하여 matching시키는 단순하고 효율적인 방법을 제안한다. 물체의 해당 면을 인식하기 위해 단순히 물체의 면을 나타내는 고유 값만을 비교하기 때문에 인식에 따르는 불필요한 절차를 줄일 수 있다. 여기서, 사용되는 물체의 면을 나타내는 고유값(SRV)은 그 물체의 view-point에 독립적인 특징들에 대하여 가중치를 부여한 값들의 합으로 얻어진다. 따라서, 이 방법은 단순하고 효율적이며 센서에 얻어진 데이터의 잡음과 occlusion에 덜 민감하다.