• Proceedings of the Earthquake Engineering Society of Korea Conference
• /
• 2006.03a
• /
• pp.463-470
• /
• 2006

이 논문에서는 신경망 제어기와 MR 댐퍼를 이용하여 지진하중을 받는 지진격리 벤치마크 구조물의 응답 감소를 위한 반능동 제어방법이 제안되었다. 제안방법 중 신경망 제어기에는 적절한 제어력을 산출하기 위해 가격함수를 기반으로한 학습 알고리즘과 간편한 민감도 계산기법이 도입되었다. MR 댐퍼가 계산되어진 제어력과 비슷한 제어력을 낼 수 있도록 clipped 알고리즘을 이용하였고, 제안된 반능동 신경망 제어방법이 지진격리 장치가 설치된 벤치마크 구조물에 적용되었다. 수치해석에서는 벤치마크 문제를 정의한 논문에서 제공된 수동제어방법이나 예시제어방법과 제안 방법의 제어성능을 비교하였다. 수치해석 결과 제안방법은 지하 변위를 약간 증가시키지만, 각층의 가속도, base shear, building corner drift 등을 매우 효과적으로 줄이는 것으로 판명되었다.

• The Journal of Korean Institute of Electromagnetic Engineering and Science
• /
• v.26 no.11
• /
• pp.1012-1019
• /
• 2015

This paper shows the acceleration of iterative physical optics(IPO) for radar cross section(RCS) by using two techniques effectively. For the analysis of the multiple reflection in the cavity, IPO uses the near field method, unlike shooting and bouncing rays method which uses the geometric optics(GO). However, it is still far slower than physical optics(PO) and it is needed to accelerate the speed of IPO for practical purpose. In order to address this problem, graphic processing unit(GPU) can be applied to reduce calculation time and adaptive iterative physical optics-change rate(AIPO-CR) method is also applicable effectively to optimize iteration for acceleration of calculation.

• Proceedings of the Korean Institute of Intelligent Systems Conference
• /
• 2006.11a
• /
• pp.297-300
• /
• 2006

본 논문에서는 비선형 기동표적의 추적에 대한 새로운 접근 방식을 소개한다. 이 논문에서는 표적의 가속도를 시변 변수인 표적의 추가적인 잡음으로 두고 각각의 가속도 간격의 정도에 따라 얻어지는 모든 잡음에 대한 변수에 의해 각각의 하부 모델들을 특성화시켰다. 표적의 기동중에 나타나는 가속도를 효과적으로 다루기 위하여, 잡음의 크기가 급격히 증가할 경우 증가분을 가속도로 인식하여 기동표적 관계식에 이용하였다. 또한 모르는 가속도에 따른 시변 변수를 적응적으로 어립잡기는 어렵기 때문에 정밀한 계산을 위하여 퍼지 뉴럴 네트워크와 적응 상호작용 다중모델 기법을 이용하였다. 퍼지 뉴럴 네트워크의 동정을 위해서는 오차 역전파 학습법을 사용하였다. 그리고 제안된 알고리즘의 수행 가능성을 보여주기 위하여 몇 가지 예를 제시하였다.

• Journal of Navigation and Port Research
• /
• v.32 no.3
• /
• pp.215-221
• /
• 2008

Dynamic characteristics of a structure, i.e., natural frequency and mode shape, have been widely using as an input data in the area of structural integrity or health monitoring which combined with the damage evaluation and structural system identification techniques. It is very difficult, however, to get those information by the conventional modal analysis method from large structures, such as the offshore structure or the long-span bridge, since the source of vibration is not available. In this paper, a method to obtain the frequencies and the mode shapes of a large span truss structure using only acceleration responses is studied. The calculation procedures to obtain acceleration responses and frequency response functions are provided utilizing a numerical model of the truss, and the process to extract natural frequencies and mode shapes from the modal analysis is cleary explained. The extracted mode shapes by proposed method are compared with those from eigenvalue analysis for the estimation of accuracy. The validity of the mode shapes is also demonstrated using an existing damage detection technique for the truss structure by simulated damage cases.

• Proceedings of the KIEE Conference
• /
• 2011.07a
• /
• pp.1794-1795
• /
• 2011

본 연구에서는 가속도 센서의 값을 디지털 신호 처리 과정을 통하여 저역통과 필터(low pass filter), 벡터의 크기(vector magnitude), 롤(roll) 그리고 피치(pitch)를 계산하는 알고리즘을 적용하였다. 필터의 경우 IIR(Infinite Impulse Response)을 이용하였으며 차수는 9차로 하였다. 피험자의 연령은 $25{\pm}5$세의 10명을 기준으로 실험하였으며 앞, 뒤, 좌, 우 방향으로 직각 낙하하도록 하였고 센서 모듈은 오른쪽 허리의 정중앙에 착용하도록 하여 피험자간의 오차가 발생하지 않도록 하였다. 환자의 낙상을 검출하기 위해서 벡터의 크기를 사용하였고 롤과 피치를 이용하여 환자의 낙상 방향을 검출하였다. 결과적으로 피험자 10명의 경우 낙상의 검출률은 100% 였으며 낙상 방향에 따른 앞, 뒤, 좌, 우 판별 정확도는 95% 정도이다. 낙상 방향의 판별은 사고 후 환자를 다룰 때의 주의할 신체부위를 참고하며 재활 운동 시 하체의 어느 쪽이 낙상의 주요인인지 분석하는 보조 자료가 될 수 있다.

• Proceedings of the Korean Society of Propulsion Engineers Conference
• /
• 2011.11a
• /
• pp.263-267
• /
• 2011

This paper presents a numerical investigation of the deflagration to detonation transition (DDT) of flame acceleration by a shock wave filled with an ethylene-air mixture in bent tube. A model consisting of the reactive compressible Navier-Stokes equations and the ghost fluid method (GFM) for complex boundary treatment is used. A various intensities of incident shock wave simulations show the generation of hot spots by shock-flame interaction and the accelerated flame propagation due to geometrical effect. Also the first detonation occurs nearly constant chemical heat release rate, 20 MJ/($g{\cdot}s$). Through our simulation's results, we concentrate the complex confinement effects in generating strong shock wave, shock-flame interaction, hot spot and DDT in pipe.

• Journal of the Korean Institute of Intelligent Systems
• /
• v.16 no.4
• /
• pp.472-478
• /
• 2006

This paper presents a new fuzzy-neural-network based interacting multiple model (FNNBIMM) algorithm for tracking a maneuvering target. To effectively handle the unknown target acceleration, this paper regards it as additional noise, time-varying variance to target model. Each sub model characterized by the variance of the overall process noise, which is obtained on the basis of each acceleration interval. Since it is hard to approximate this time-varying variance adaptively owing to the unknown acceleration, the FNN is utilized to precisely approximate this time-varying variance. The error back-propagation method is utilized to optimize each FNN. To show the feasibility of the proposed algorithm, a numerical example is provided.

• Journal of the Society of Naval Architects of Korea
• /
• v.37 no.1
• /
• pp.67-81
• /
• 2000

An accurate and efficient numerical method for two-dimensional nonlinear radiation problem has been developed. The wave motion due to a moving body is described by the assumption of ideal fluid flow, and the governing Laplace equation can be effectively solved by the higher-order boundary element method with the help of the GMRES (Generalized Minimal RESidual) algorithm. The intersection or corner problem is resolved by utilizing the so-called discontinuous elements. The implicit trapezoidal rule is used in updating solutions at new time steps by considering stability and accuracy. Traveling waves caused by the oscillating body are absorbed downstream by the damping zone technique. It is demonstrated that the present method for time marching and radiation condition works efficiently for nonlinear radiation problem. To avoid the numerical instability enhanced by the local gathering of grid points, the regriding technique is employed so that all the grids on the free surface may be distributed with an equal distance. This makes it possible to reduce time interval and improve numerical stability. Special attention is paid to the local flow around the body during time integration. The nonlinear radiation force is calculated by the "acceleration potential technique". Present results show good agreement with other numerical computations and experiments.

• The Journal of the Korea Contents Association
• /
• v.13 no.4
• /
• pp.35-43
• /
• 2013

Exercise such as walking is helpful to manage one's own weight and to counter life habit diseases such as obesity. Calorie consumption is usually calculated based on the distance walked. One way to measure the distance is by using steps and stride length. Most pedometers, including some applications in smart devices, are inaccurate, because they use a common value as the average stride length, even though each person has a different stride length. Moreover, the stride length differs depending on the walking pace, which will further increase the error. To address this, in this paper, I classify paces into three categories. Following that, I introduce a customized measurement of stride length, which is calculated based on the stride length corresponding to each pace category after obtaining x, y, z values from a 3-axis accelerometer in the smart device. In addition to this, I developed an application running on the smart device designed for the proposed measurement of stride. I have conducted three experiments for the assessment of the proposed measurement. In conclusion, I confirmed the effectiveness of this system.

• Journal of the Korea Institute of Information Security & Cryptology
• /
• v.33 no.6
• /
• pp.919-928
• /
• 2023

Deep learning technology is used in various fields such as self-driving cars, image creation, and virtual voice implementation, and deep learning accelerators have been developed for high-speed operation in hardware devices. However, several side channel attacks that recover secret information inside the accelerator using side-channel information generated when the deep learning accelerator operates have been recently researched. In this paper, we implemented a DNN(Deep Neural Network)-based MNIST digit classifier on a microprocessor and attempted a correlation power analysis attack to confirm that the weights of deep learning accelerator could be sufficiently recovered. In addition, to counter these power analysis attacks, we proposed a Node-CUT shuffling method that applies the principle of misalignment at the time of power measurement. It was confirmed through experiments that the proposed countermeasure can effectively defend against side-channel attacks, and that the additional calculation amount is reduced by more than 1/3 compared to using the Fisher-Yates shuffling method.