• 한국통신학회논문지
• /
• 제37B권11호
• /
• pp.1022-1032
• /
• 2012

본 논문에서는 OFDMA 기반의 무선 메쉬 네트워크에서 다수의 노드 간 발생하게 되는 주파수 비동기 문제를 해결하기 위해 생체모방 알고리즘(bio-inspired algorithm)을 이용하여 인접 노드 간 지역적인 주파수 동기화를 통해 메쉬 네트워크 전체를 하나의 주파수로 수렴시켜 나가는 분산 주파수 동기화 방식을 제안한다. 메쉬 네트워크의 주파수 수렴 특성은 네트워크의 규모와 구성 노드들의 배치에 따라 서로 다르기 때문에 특정 토폴로지의 경우 주파수 수렴을 위해 많은 시간이 소요될 수 있다. 제안하는 기법은 가중치 적용을 통하여 메쉬 토폴로지 형태에 크게 의존하지 않는 빠른 주파수 동기화를 이룰 수 있음을 확인한다.

• 한국전자파학회논문지
• /
• 제25권7호
• /
• pp.757-766
• /
• 2014

본 논문은 거시적인 접근방법을 이용하여 고출력 전자기파에 대한 피해 비용을 보다 실제적으로 산출하기 위한 방법을 제시하고, 이에 대한 피해액을 계산하였다. 먼저 취약성 분석을 통하여 사회 기반 시설 중 사회적으로 가장 큰 혼란을 야기할 수 있을 것이라고 판단되는 전력망에 집중하여 연구하였으며, 이에 대해 고출력 전자기파에 의한 피해액 계산식은 GDP에 대한 총 에너지 소비를 고려하여 고출력 전자기파 피해 시 발생할 수 있는 피해 손실을 효과적으로 예측하기 위한 방법을 제시하였다. 피해액은 고출력 전자기파의 공격 형태에 따라 피해 범위를 설정하고 계산하였으며, 이러한 피해액 계산을 통하여 고출력 전자기파에 취약한 시설에 대해서는 피해액 산출 후 방호 비용과 비교하여 시설 방호 대책을 수립할 수 있는 기반을 마련할 수 있을 것이다.

• 대한전기학회:학술대회논문집
• /
• 대한전기학회 1999년도 하계학술대회 논문집 G
• /
• pp.3015-3017
• /
• 1999

KEPCO is faced with a number of network choice for Distribution Automation System(DAS), such as Power Line, Pair. Coaxial and Optical Cable, etc. The increasing use of DAS requires suitable communication networks. KEPCO has exhausted much efforts to implement its own standard DAS in possible early date. This paper presents the DAS in KEPCO and some of initial design efforts toward KEPCO's DAS at KEPRI.

• 한국원자력학회:학술대회논문집
• /
• 한국원자력학회 1997년도 추계학술발표회논문집(1)
• /
• pp.58-63
• /
• 1997

This paper is concerned with an algorithm based on neural networks to calculate the axial power distribution using excore defector signals in the nuclear reactor core. The fundamental basis of the algorithm is that the detector response can be fairly accurately estimated using computational codes. In other words, the training set, which represents relationship between detector signals and axial power distributions, for the neural network can be obtained through calculations instead of measurements. Application of the new method to the Yonggwang nuclear power plant unit 3 (YGN-3) shows that it is superior to the current algorithm in place.

• 대한전자공학회:학술대회논문집
• /
• 대한전자공학회 2007년도 하계종합학술대회 논문집
• /
• pp.171-172
• /
• 2007

The need for open architecture based IP network is becoming increasingly critical because the electric power industry has begun to upgrade to digital systems. In this paper, we implemented IPv6 testbed and experimented IPv6 performance actually for examination on IPv6 applications to electric power communication network. We achieved preliminary tests relevant to IPv6 to solve expected problems before.

• 대한전기학회:학술대회논문집
• /
• 대한전기학회 2000년도 하계학술대회 논문집 A
• /
• pp.182-184
• /
• 2000

There are several facilities in Kochang distribution test center such as artificial fault generator(AFG), new distribution automation system(NDAS), communication networks (wireless and optic), lumped constant circuit, switches for distribution automation, overhead and underground distribution line. We have been field testing on remote control, data acquisition. remote metering, feeder automation and so on for distribution automation using those equipment.

• 한국전자파학회:학술대회논문집
• /
• 한국전자파학회 2002년도 종합학술발표회 논문집 Vol.12 No.1
• /
• pp.366-369
• /
• 2002

Two artificial neural networks (ANN) are used to model the electric coupling probe in the combline resonators. One is used to analyze and synthesize the electric probe, and the other is used to correct errors between the results of the analysis and the synthesis ANNs and the fabrication results. The ANNs for the analysis and the synthesis of the electric probe are trained using the physical dimensions of the electric probe and the corresponding coupling bandwidth which is obtained using the finite element method. The ANNs for the error correction are trained using a very small set of the measurement results. Once trained, the ANN models provide the correct result approaching the accuracy of the measurement. The results from the ANN models show fairly good agreement with those of the measurement and they can be used as good initial design values.

• 한국정보통신학회논문지
• /
• 제19권8호
• /
• pp.1917-1924
• /
• 2015

무선 센서 네트워크에서 멀리떨어진 목적노드까지 정보를 전달하기 위해서는 멀티홉전송을 통한 효율적인 라우팅기법이 필요로 된다. 멀티홉전송에 있어서 다중경로 라우팅기법을 사용하게 되면 특정 경로가 사용 불능 상태가 되거나 트래픽이 크게 증가하는 경우에도 다중경로를 활용한 안정적 데이터 전송이 가능하게 된다. 본 논문에서는 경로 전체에 대해 다중경로를 사용하는 기존 다중경로 라우팅 기법을 개선하여 일부 열악한 링크 구간만을 다중경로로 전송하는 부분 다중경로라우팅 방식을 제안함으로써 안전하고 빠른 데이터 전송을 보장함과 동시에 불필요하게 전송에 참여하는 노드의 수를 최소화함으로써 노드의 전력소모를 최소화하고 네트워크를 효율적으로 사용할 수 있도록 하였다.

• KSII Transactions on Internet and Information Systems (TIIS)
• /
• 제15권12호
• /
• pp.4326-4344
• /
• 2021

The SinGAN is one of generative adversarial networks that can be trained on a single nature image. It has poor ability to learn more global features from nature image, and losses much local detail information when it generates arbitrary size image sample. To solve the problem, a non-linear function is firstly proposed to control downsampling ratio that is ratio between the size of current image and the size of next downsampled image, to increase the ratio with increase of the number of downsampling. This makes the low-resolution images obtained by downsampling have higher proportion in all downsampled images. The low-resolution images usually contain much global information. Therefore, it can help the model to learn more global feature information from downsampled images. Secondly, the attention mechanism is introduced to the generative network to increase the weight of effective image information. This can make the network learn more local details. Besides, in order to make the output image more natural, the TVLoss function is introduced to the loss function of SinGAN, to reduce the difference between adjacent pixels and smear phenomenon for the output image. A large number of experimental results show that our proposed model has better performance than other methods in generating random samples with fixed size and arbitrary size, image harmonization and editing.

• Transactions on Control, Automation and Systems Engineering
• /
• 제4권2호
• /
• pp.124-129
• /
• 2002

This paper demonstrates that the largest Lyapunov exponent λ of recurrent neural networks can be controlled efficiently by a stochastic gradient method. An essential core of the proposed method is a novel stochastic approximate formulation of the Lyapunov exponent λ as a function of the network parameters such as connection weights and thresholds of neural activation functions. By a gradient method, a direct calculation to minimize a square error (λ - λ$\^$obj/)$^2$, where λ$\^$obj/ is a desired exponent value, needs gradients collection through time which are given by a recursive calculation from past to present values. The collection is computationally expensive and causes unstable control of the exponent for networks with chaotic dynamics because of chaotic instability. The stochastic formulation derived in this paper gives us an approximation of the gradients collection in a fashion without the recursive calculation. This approximation can realize not only a faster calculation of the gradient, but also stable control for chaotic dynamics. Due to the non-recursive calculation. without respect to the time evolutions, the running times of this approximation grow only about as N$^2$ compared to as N$\^$5/T that is of the direct calculation method. It is also shown by simulation studies that the approximation is a robust formulation for the network size and that proposed method can control the chaos dynamics in recurrent neural networks efficiently.