• 전기전자학회논문지
• /
• 제16권4호
• /
• pp.328-334
• /
• 2012

본 논문에서는 레이더 단면적(RCS, Radar Cross Section)을 감소시키기 위한 전자파 구조로서 십자형 다이폴 슬롯을 원소로 가지며 다른 배열형태 구조를 갖는 구형 주파수 선택적 표면(FSS, Frequency Selective Surface) 구조의 전파특성을 해석하였다. 유한크기와 곡면형상을 갖는 구형 주파수 선택적 표면 구조의 주파수 특성을 해석하기 위해 RWG 함수를 적용한 3차원 모멘트법을 사용하였으며 반복법 중 하나인 BiCGSTab(Biconjugate Gradient Stabilized) 알고리즘을 적용하여 해석시간의 효율성을 개선하였다. 1m 직경을 갖는 완전도체(PEC, Perfect electric conductor)의 구에 대한 이론적 해석결과인 Mie의 RCS 특성과 비교 검증하여 제안한 해석 알고리즘의 유효함을 입증하였다. 구형 주파수 선택적 표면 구조에서 슬롯 성분의 배열 방법에 따른 RCS 변화를 관찰하여 비교하였으며, 곡면형상의 주파수 선택구조 설계시 배열의 방법도 RCS 특성에 중요한 변수가 될 수 있음을 확인하였다.

• 한국멀티미디어학회논문지
• /
• 제12권7호
• /
• pp.997-1005
• /
• 2009

본 논문에서는 계층간 설계 방식에 근거하여 에너지 효율적인 센서 네트워크 클러스터링 기법을 제안한다. 제안된 기법은 무선 센서 네트워크의 응용 환경 특성에 적합하도록 동작한다. 본 논문에서는 응용계층 정보를 이용하여 이벤트가 발생할 경우 이벤트가 발생한 지역의 센서 노드들로 구성된 클러스터와 그 이외 지역의 클러스터들로 구성하는 클러스터링 기법을 제안하였다. 제안된 클러스터링 기법에서는 여러 클러스터를 경유하여 다중 경로로 이벤트를 전달하는데 따른 에너지 소모를 절감할 수 있다. 또한 이벤트와 무관한 클러스터에서 에너지를 절약하기 위하여 각 노드 당 한 개의 슬롯만을 할당하는 TDMA 스케줄링 기법을 제안하였다. 제안하는 클러스터링 기법은 전체 네트워크의 수명을 증가시킬 수 있으며 이벤트의 발생 주기, 지속 시간, 범위에 따른 시뮬레이션을 통하여 에너지 효율성 관점에서 우수함을 입증하였다.

• International Journal of Ocean System Engineering
• /
• 제2권1호
• /
• pp.50-56
• /
• 2012

This study investigates flow fields and energy dissipation due to regular wave interaction with a perforated vertical breakwater, through velocity data measurement in a two-dimensional wave tank. As the waves propagate through the perforated breakwater, the incoming wave energy is reflected back to the ocean, dissipated due to very turbulent flows near the perforations and inside the chamber, and transmitted through the perforations of the breakwater. This transmitted energy is further reduced due to the presence of the perforated back wall. Hence most of the energy is either reflected or dissipated in the vicinity of the structure, and only a small amount of the incoming wave energy is transmitted through the structure. In this study, particle image velocimetry (PIV) technique was employed to measure two-dimensional instantaneous velocity fields in the vicinity of the structure. Measured velocity data was treated statistically, and used to calculate mean flow fields, turbulence intensity and turbulent kinetic energy. For investigation of the flow pattern, time-averaged mean velocity fields were examined, and discussed using the cross-sections through slot and wall for comparison. Flow fields were obtained and compared for various cases with different regular wave conditions. In addition, turbulent kinetic energy was estimated as an approach to understand energy dissipation near the perforated breakwater. The turbulent kinetic energy was distributed against wave height and wave period to see the dependence on wave conditions.

• 전기학회논문지
• /
• 제67권1호
• /
• pp.114-123
• /
• 2018

This paper concerns a design and learning method of softmax function neural networks based on K-means clustering. The partial discharge data Information is preliminarily processed through simulation using an Epoxy Mica Coupling sensor and an internal Phase Resolved Partial Discharge Analysis algorithm. The obtained information is processed according to the characteristics of the pattern using a Motor Insulation Monitoring System program. At this time, the processed data are total 4 types that void discharge, corona discharge, surface discharge and slot discharge. The partial discharge data with high dimensional input variables are secondarily processed by principal component analysis method and reduced with keeping the characteristics of pattern as low dimensional input variables. And therefore, the pattern classifier processing speed exhibits improved effects. In addition, in the process of extracting the partial discharge data through the MIMS program, the magnitude of amplitude is divided into the maximum value and the average value, and two pattern characteristics are set and compared and analyzed. In the first half of the proposed partial discharge pattern classifier, the input and hidden layers are classified by using the K-means clustering method and the output of the hidden layer is obtained. In the latter part, the cross entropy error function is used for parameter learning between the hidden layer and the output layer. The final output layer is output as a normalized probability value between 0 and 1 using the softmax function. The advantage of using the softmax function is that it allows access and application of multiple class problems and stochastic interpretation. First of all, there is an advantage that one output value affects the remaining output value and its accompanying learning is accelerated. Also, to solve the overfitting problem, L2-normalization is applied. To prove the superiority of the proposed pattern classifier, we compare and analyze the classification rate with conventional radial basis function neural networks.