• KIPS Transactions on Software and Data Engineering
• /
• v.7 no.12
• /
• pp.485-496
• /
• 2018

So far, the deep learning, a field of artificial intelligence, has achieved remarkable results in solving problems from unstructured data. However, it is difficult to comprehensively judge situations like humans, and did not reach the level of intelligence that deduced their relations and predicted the next situation. Recently, deep neural networks show that artificial intelligence can possess powerful relational reasoning that is core intellectual ability of human being. In this paper, to analyze and observe the performance of Relation Networks (RN) among the neural networks for relational reasoning, two types of RN-based deep neural network models were constructed and compared with the baseline model. One is a visual question answering RN model using Sort-of-CLEVR and the other is a text-based question answering RN model using bAbI task. In order to maximize the performance of the RN-based model, various performance improvement experiments such as hyper parameters tuning have been proposed and performed. The effectiveness of the proposed performance improvement methods has been verified by applying to the visual QA RN model and the text-based QA RN model, and the new domain model using the dialogue-based LL dataset. As a result of the various experiments, it is found that the initial learning rate is a key factor in determining the performance of the model in both types of RN models. We have observed that the optimal initial learning rate setting found by the proposed random search method can improve the performance of the model up to 99.8%.

• The Journal of The Korea Institute of Intelligent Transport Systems
• /
• v.13 no.1
• /
• pp.67-74
• /
• 2014

In this paper, we designed a multiband monopole antenna for next-generation WLAN system. In conventional WLAN system, UWB antennas were used together, and, because the radiation occurs in different parts depending on the antenna structure, it has the disadvantage of having an unstable impulse response characteristic due to dispersion characteristics. Although a UWB antenna that has suitable radiation pattern for WLAN band, it does not have good impedance matching and has severe echo. Therefore, in this paper, a monopole antenna was designed by using CPW power feed so that various impedances can be easily implemented when designing an antenna and more parameters can be derived that can be used for design for optimal performance.

• Journal of Institute of Control, Robotics and Systems
• /
• v.3 no.2
• /
• pp.149-155
• /
• 1997

본 논문에서는 비선형 시변 시스템을 제어할 경우 제어시스템의 안정성을 보장하고 성능을 향상시키기 위한 새로운 적응제어 구조를 전개하였다. 주어진 플랜트가 선형 시불변이라는 가정하에 표준 기준 모델 적응제어기가 적용될 경우 발생되는 출력오차는 플랜트의 비선형 시변특성으로 인하여 점근적으로 0에 수렴되지 않는다. 이때 미지의 출력오차를 점근적으로 0에 수렴시키는 방법으로 퍼지보상기를 사용하였으며 결과적으로 플랜트의 비선형 시변 특성을 보상하는 효과를 얻을 수 있었다. 퍼지 보상기로는 출력오차등의 조건에 따라 이득이 변하는 퍼지 PID 보상기를 도입하여 안정하게 설계되도록 노력하였다. 또한 출력오차를 점근적으로 0에 수렴시키는 것은 표준 기준 모델 적응제어기 내부의 모든 파라미터와 신호가 유한하게 됨을 의미하기 때문에, 제어시스템 전체의 안정도를 보장할 뿐만 아니라 결과적으로 과도응답 성능을 향상시킬 수 있게 되었다. 몇가지 예제를 대상으로 시뮬레이션을 수행하고 그 결과를 분석함으로써 비선형 시변 시스템을 제어할 경우 본 논문에서 전개된 새로운 적응제어 구조의 타당성을 확인하였다.

• Transactions of the Korean Society of Automotive Engineers
• /
• v.14 no.6
• /
• pp.143-150
• /
• 2006

In this paper, a new bushing model for vehicle dynamics analysis using Bouc-Wen hysteretic model is proposed. Bushing components of a vehicle suspension system are tested to capture the nonlinear behavior of rubber bushing elements using the MTS 3-axes rubber test machine. The results of the tests are used to define parameters in Bouc-Wen bushing model, which was employed to represent the hysteretic characteristics of the bushing. Bushing parameters are obtained by using genetic algorithms and sensitivity analysis of parameters are also carried out. ADAMS program was used for the identification process and VisualDOC program was employed to find the optimal parameters. A half-car simulation was carried out to show the usefulness of the developed bushing model.

• Journal of the Institute of Electronics Engineers of Korea SP
• /
• v.43 no.2 s.308
• /
• pp.87-92
• /
• 2006

In array signal processing, high order statistics can be used to estimate parameters from signal of sums of complex exponential. This paper presents a high order Matrix Pencil method(MPM) using the fourth order cumulant. Since the fourth order cumulant can suppress the Gaussian noise, the response of MPM has better noise immunity than the conventional approaches. We successfully formulate the high order MPM with all the benefits of MPM along with higher accuracy. In the numerical simulations we demonstrated that the proposed method with forth order cumulant has better resolution to find degree of arrival(DOA) in the presence of the Gaussian noise.

• The Journal of the Korea institute of electronic communication sciences
• /
• v.15 no.5
• /
• pp.949-958
• /
• 2020

The functions of shock absorbers are to dampen body, suspend motions, dissipate impact energy, and control tire force variation. During the operation, hydraulic oil is passed between the chambers via a flow restrictions. Therefore the damping force characteristics of shock absorber is determined by the characteristics of orifices and flow restrictions. The uncertainty in design variable affects the performance of suspension system strongly. But, the researches about the influence of uncertainty in design variable such as a fluid restriction's property of shock absorber, on the suspension system performance was hardly ever proposed. In this paper, we used statistical method of Latin Hypercube sampling, and the effects of design variables uncertainty on the performance of suspension system was presented.

• Analytical Science and Technology
• /
• v.7 no.3
• /
• pp.349-359
• /
• 1994

A gas chromatographic method for analyzing the gas odorants concentration in natural gas was studied. Eight odorants involving TBM and THT were completely separated by using OV-10 column. The optimization of several interrelated key parameters affecting the response of FPD such as hydrogen flow rate, air flow rate and detector temperature were accomplished. A permeation device was used to obtain calibration curves of TBM and THT. This analytical method has applied to measure TBM and THT used as a natural gas odorant blend in natural gas pipeline. In order to elucidate the relationship between odor level and odorant level feasibility test of fragrance meter was demonstrated.

• Journal of the Korean Society for Aeronautical & Space Sciences
• /
• v.38 no.2
• /
• pp.188-194
• /
• 2010

In this paper, the typical Synthetic Aperture Radar (SAR) image quality parameters and analysis method are defined, and the SAR image analysis tool is presented for SAR image evaluation. The structure of the developed SAR image analysis tool consists of four key modules; point target analysis (PTA) module, distributed target analysis (DTA) module, ambiguity analysis (AMA) module, and NESZ analysis (NESZA) module. The developed tool is able to extract the various SAR system parameters from standard SAR product format files. Based on these extracted system parameters, typical SAR image quality parameters are derived from SAR image data.

• Journal of Satellite, Information and Communications
• /
• v.9 no.1
• /
• pp.38-44
• /
• 2014

In this paper, we designed a multiband monopole antenna for next-generation WLAN system. In conventional WLAN system, UWB antennas were used together, and, because the radiation occurs in different parts depending on the antenna structure, it has the disadvantage of having an unstable impulse response characteristic due to dispersion characteristics. Although a UWB antenna that has suitable radiation pattern for WLAN band, it does not have good impedance matching and has severe echo. Therefore, in this paper, a monopole antenna was designed by using CPW power feed so that various impedances can be easily implemented when designing an antenna and more parameters can be derived that can be used for design for optimal performance.

• Journal of Advanced Navigation Technology
• /
• v.13 no.5
• /
• pp.756-762
• /
• 2009

As nonlinear friction, stick-slip friction in hydraulic actuators are a problem for accuracy and repeatability. Therefore friction compensation has been approached through various control algorithms. A Adaptive discrete time sliding mode tracking controller has been applied in order to compensate the nonlinear friction characteristics in a hydraulic Actuator. Based on the diophantine equation, a new discrete time sliding function is defined and utilized for the control law which includes a friction and modeling error. Robustness is increased by using both a projection algorithm and a sliding function-based nonlinear feedforward. From the results of simulation and experiment good tracking performance is achieved.