• 조명전기설비학회논문지
• /
• 제24권8호
• /
• pp.55-61
• /
• 2010

본 논문은 클래스 내와 클래스 간의 분산정보를 이용한 주파수 성분의 선택적 추출기법과 PCA-ELM 기반의 유도전동기 고장진단 시스템을 제안한다. 제안된 방법은 취득된 전류신호를 DFT에 의해 주파수 영역으로 변환한 후 분산정보를 이용하여 고장상태별로 차별성이 큰 순서대로 주파수 성분을 추출한다. 다음 단계로 선택된 주파수 성분에 대해서 PCA를 이용하여 고장상태별 특징들을 추출한다. 마지막 단계는 학습속도가 매우 우수한 ELM분류기에 의해 유도전동기의 상태를 진단하게 된다. 다양한 부하에 대하여 몇몇의 전기적 고장과 기계적 고장 하에서 획득한 데이터를 이용하여 제안된 방법의 타당성을 검증한다.

• Geomechanics and Engineering
• /
• 제6권6호
• /
• pp.531-544
• /
• 2014

Studies of earthquakes over the last 50 years and the examination of dynamic soil behavior reveal that soil behavior is highly nonlinear and hysteretic even at small strains. Nonlinear behavior of soils during a seismic event has a predominant role in current site response analysis approaches. Common approaches to ground response analysis include linear, equivalent linear and nonlinear methods. These methods of ground response analysis may also be categorized into time domain and frequency domain concepts. Simplicity in developing analytical relations and accuracy in considering soils' dynamic properties dependency to loading frequency are benefits of frequency domain analysis. On the other hand, nonlinear methods are complicated and time consuming mainly because of their step by step integrations in time intervals. In part Ι of this paper, governing equations for seismic response analysis of surcharged and layered soils were developed using fundamental of wave propagation theory based on transfer function and boundary conditions. In this part, nonlinear seismic ground response is analyzed using extended HFTD method. The extended HFTD method benefits Newton-Raphson procedure which applies regular iterations and follows soils' fundamental stress-strain curve until convergence is achieved. The nonlinear HFTD approach developed here are applied to some examples presented in this part of the paper. Case studies are carried in which effects of some influencing parameters on the response are investigated. Results show that the current approach is sufficiently accurate, efficient, and fast converging. Discussions on the results obtained are presented throughout this part of the paper.

• 한국정보통신학회:학술대회논문집
• /
• 한국해양정보통신학회 2005년도 춘계종합학술대회
• /
• pp.273-276
• /
• 2005

This paper presents an electronic ballast using a step down converter, a low frequency inverter for high pressure discharge lamp. The proposed ballast is composed of a full-wave rectifier, a step down converter operated as a current source with power regulation and a low frequency inverter with ignition circuit. The ignition circuit generates high voltage pulse of 1${\sim}$2[kV] peak, 130[Hz]. Moreover, it is able to reignite at regular intervals by protective circuit. As experimental results on the test, lamp voltage, current and consumption power are measured 132.5[V], 7.6[A] and 1,005[W], respectively. It was confirmed that the designed ballast operate the lamp with a constant power.

• 전력전자학회:학술대회논문집
• /
• 전력전자학회 2007년도 하계학술대회 논문집
• /
• pp.403-405
• /
• 2007

This paper presents the over-modulation strategy and indirect vector control drive of NPC type PWM inverter. NPC inverter has three level phase voltage output.It can perform in high voltage through assembling switching components. It has less harmonics and surge voltage stress at motor terminals than the 2 level inverter in same switching frequency through 3 level voltage. The conventional railway vehicle has used the vector control to MI=0.907 and the slip-frequency control from MI=0.907 to six-step mode. The slip-frequency control has bad motive power and slow torque control response. But vector control has good motive power and can instant torque control. In this paper, output voltage is controlled linearly from linear region to six-step mode by using over-modulation strategy. And NPC inverter is used.

• 한국지능시스템학회논문지
• /
• 제20권3호
• /
• pp.380-387
• /
• 2010

본 논문에서는 3상 유도전동기의 고장진단을 수행하기 위해 상호정보량과 선형판별분석기법에 기반을 둔 진단 알고리즘을 제안한다. 실험 장치는 유도전동기 구동의 기계적 모듈과 고장신호를 구하기 위한 데이터 획득 모듈로 구성하였다. 제안된 방법은 취득된 전류신호를 DFT에 의해 주파수 영역으로 변환한 후 분산정보를 이용하여 고장상태별로 차별성이 큰 순서대로 유효 주파수 성분을 추출한다. 다음 단계로 선택된 주파수 성분에 대해서 선형판별분석기법을 적용하여 고장상태별 특징들을 추출한 후 k-NN 분류기에 의해 유도전동기의 상태를 진단하게 된다. 제안된 방법의 타당성을 보이기 위해 다양한 조건하에서 실험한 결과 기존방법에 비하여 우수한 결과를 나타냈다.

• International Journal of Naval Architecture and Ocean Engineering
• /
• 제13권1호
• /
• pp.223-235
• /
• 2021

The Vortex-Induced Vibration (VIV) test system on deepwater riser based on Bare Fiber Bragg Grating (BFBG) sensor technology was designed. Meanwhile, a riser VIV response numerical model was established based on the work-energy principle. The results show that the first-order vibration frequency dominates the vibration of the riser, and as the velocity increases, the dominant frequency of the riser gradually increases under the effect of different top tensions. At the same velocity, as the top tension increases step by step, the dominant frequency and fatigue damage at the same position along the axial length of the riser both gradually decreases. The model test and numerical simulation show a relatively consistent change, maintaining a high degree of agreement. The process control system based on BFBG of model test has excellent performance, and FBG sensors have great advantages in VIV test of a vertical riser in water.

• Smart Structures and Systems
• /
• 제30권4호
• /
• pp.353-369
• /
• 2022

In this paper, a new sensor chip with frequency reconstruction range of 2.252 GHz ~ 2.450 GHz is designed and fabricated. On this basis, a self-designed "T-shaped" shell is added to overcome the disadvantage of uneven deformation of the traditional steel shell, and the range of the sensor chip is expanded to 0 kN ~ 96 kN. The liquid metal antenna is used to carry out a step-by-step loading test, and the relationship between the antenna resonance frequency and the pressure load is analyzed. The results show that there is a good linear relationship between the pressure load and the resonant frequency. Therefore, the liquid metal antenna can be regarded as a pressure sensor. The cyclic loading and unloading experiments of the sensor are carried out, and different loading rates are used to explore the influence on the performance of the sensor. The loading and unloading characteristic curves and the influence characteristic curves of loading rate are plotted. The experimental results show that the sensor has no residual deformation during the cycle of loading and unloading. Moreover, the influence of temperature on the performance of the sensor is studied, and the temperature correction formula is derived.

• 전기학회논문지
• /
• 제57권11호
• /
• pp.2041-2046
• /
• 2008

A frequency-dependent slow-wave factor (SWF) and equivalent circuit model of transmission line with defected ground structures (DGS) is described. Once S-parameters of a DGS transmission line are given, the conventional frequency -independent equivalent circuit elements are extracted using 3dB cutoff and resonant frequencies (Fc and Fo) as the first step. Using the initial equivalent elements and simple transmission line theories, a frequency-dependent equivalent transmission line model is established through an analytical method, and finally the frequency dependent SWF is calculated. The proposed equivalent circuit model and SWF are frequency-dependent and more reliable because even small insertion loss within available passband is considered, while they have been independent of frequency.

• 지구물리와물리탐사
• /
• 제16권3호
• /
• pp.139-144
• /
• 2013

유한길이의 다중 송수신 쌍극자에 의한 수평다층구조의 시간영역 전자기장을 계산하기 위한 컴퓨터 프로그램을 개발하였다. 시간영역 반응은 주파수영역에서 계산된 값에 빠른 역푸리에변환(inverse fast Fourier transform: FFT)을 적용하여 효율적으로 얻을 수 있다. 먼저 대수영역에서 등간격으로 한 decade 당 10개의 주파수영역 반응을 구한 후 FFT를 적용시키기 위해 3차 스플라인 사이채움(cubic spline interpolation)을 실시한다. 이 때 위상의 경우에는 스플라인 사이채움 이전에 위상곡선을 연속적으로 만들어 주는 과정이 추가된다. 스플라인 사이채움된 자료들은 송신전류파형과 곱말기(convolution)를 한 후 FFT를 통해 시간영역 자료로 만들어진다. 이 논문에서는 step-off 파형만 고려하였다. 개발된 시간영역 프로그램은 해석해와 해양 탄화수소 저류층 모델에 대한 반응을 이용하여 검증하였으며, 그 결과는 충분히 정확함을 확인 할 수 있었다.

• 한국수학교육학회지시리즈A:수학교육
• /
• 제49권1호
• /
• pp.111-118
• /
• 2010

This paper attempts to establish a scoring system for the originality in evaluation of mathematical creativity. The scoring system is composed of three categories; fluency, flexibility and originality. In this paper, we proposed an evaluation method for originality as following based on relative frequency and standard normal distribution. (1) Fluency: It is judged on the basis of the number of correct answers a student made. If several correct answers are given for a single category, then its maximum score is set to 5 points. (2) Flexibility: We examined how many categories the students' responses can be classified into. If at most 15 answers are allowed for each question, the maximum score of flexibility is 15 points. (3) Originality: Originality score is given if a student made some original response that other students did not show. That is, it reflects relative rarity. The originality is measured according to the following steps: Step 1: Analyze the frequency of how many students made an answer to the response type categorized at low level, and calculate the relative frequency p of each category. Step 2: Find the originality point os for each response, that is, os = max{0,z} where z satisfies P(Z > z) = p with standard normal distributed random variable Z. For example, - p is greater than 0.5: 0 point - p is 0.1587: 1 point - p is 0.0228: 2 points - p is 0.0013: 3 points Step 3: Assign the one's originality score to the sum of originality point for each response. Remark. There is no upper limit of originality score.