• 제목/요약/키워드: dynamic tunneling

검색결과 39건 처리시간 0.024초

Compression of Image Data Using Neural Networks based on Conjugate Gradient Algorithm and Dynamic Tunneling System

  • Cho, Yong-Hyun;Kim, Weon-Ook;Bang, Man-Sik;Kim, Young-il
    • 한국지능시스템학회:학술대회논문집
    • /
    • 한국퍼지및지능시스템학회 1998년도 The Third Asian Fuzzy Systems Symposium
    • /
    • pp.740-749
    • /
    • 1998
  • This paper proposes compression of image data using neural networks based on conjugate gradient method and dynamic tunneling system. The conjugate gradient method is applied for high speed optimization .The dynamic tunneling algorithms, which is the deterministic method with tunneling phenomenon, is applied for global optimization. Converging to the local minima by using the conjugate gradient method, the new initial point for escaping the local minima is estimated by dynamic tunneling system. The proposed method has been applied the image data compression of 12 ${\times}$12 pixels. The simulation results shows the proposed networks has better learning performance , in comparison with that using the conventional BP as learning algorithm.

L형 터널 트랜지스터의 트랩-보조-터널링 현상 조사 (Investigation of Trap-Assisted-Tunneling Mechanism in L-Shaped Tunneling Field-Effect-Transistor)

  • 파라즈 나잠;유윤섭
    • 한국정보통신학회:학술대회논문집
    • /
    • 한국정보통신학회 2018년도 추계학술대회
    • /
    • pp.512-513
    • /
    • 2018
  • 트랩-보조-터널링(Trap-Assisted-Tunneling; TAT)은 실제 터널링 전계 효과 트랜지스터 (TFET)의 임계 이하 기울기를 저하시키고 시뮬레이션에서 고려되어야한다. 그러나, 그 메커니즘은 라인 터널링 타입 L형 TFET(LTFET)에서는 잘 알려져 있지 않았다. 본 연구는 dynamic nonlocal Schenk 모델을 이용한 LTFET의 TAT 메커니즘을 연구한다. 이 연구에서는 터널링 이벤트를 위해서 phonon assisted and direct band가 모두 고려되었다.

  • PDF

후향전파 알고리즘과 동적터널링 시스템을 조합한 다층신경망의 새로운 학습방법 (A new training method of multilayer neural networks using a hybrid of backpropagation algorithm and dynamic tunneling system)

  • 조용현
    • 전자공학회논문지B
    • /
    • 제33B권4호
    • /
    • pp.201-208
    • /
    • 1996
  • This paper proposes an efficient method for improving the training performance of the neural network using a hybrid of backpropagation algorithm and dynamic tunneling system.The backpropagation algorithm, which is the fast gradient descent method, is applied for high-speed optimization. The dynamic tunneling system, which is the deterministic method iwth a tunneling phenomenone, is applied for blobal optimization. Converging to the local minima by using the backpropagation algorithm, the approximate initial point for escaping the local minima is estimated by the pattern classification, and the simulation results show that the performance of proposed method is superior th that of backpropagation algorithm with randomized initial point settings.

  • PDF

최적화용 신경망의 성능개선을 위한 새로운 최적화 기법 (A new optimization method for improving the performance of neural networks for optimization)

  • 조영현
    • 전자공학회논문지C
    • /
    • 제34C권12호
    • /
    • pp.61-69
    • /
    • 1997
  • This paper proposes a new method for improving the performances of the neural network for optimization using a hyubrid of gradient descent method and dynamic tunneling system. The update rule of gradient descent method, which has the fast convergence characteristic, is applied for high-speed optimization. The update rule of dynamic tunneling system, which is the deterministic method with a tunneling phenomenon, is applied for global optimization. Having converged to the for escaping the local minima by applying the dynamic tunneling system. The proposed method has been applied to the travelling salesman problems and the optimal task partition problems to evaluate to that of hopfield model using the update rule of gradient descent method.

  • PDF

하이브리드 알고리즘을 이용한 신경망의 학습성능 개선 (Improving the Training Performance of Neural Networks by using Hybrid Algorithm)

  • 김원욱;조용현;김영일;강인구
    • 한국정보처리학회논문지
    • /
    • 제4권11호
    • /
    • pp.2769-2779
    • /
    • 1997
  • 본 논문에서는 공액기울기법과 터널링 시스템을 조합사용하여 신경망의 학습성능을 향상시킬 수 있는 효율적인 방법을 제안하였다. 빠른 수렴속도의 학습을 위하여 공액 기울기법에 기초한 후향전파 알고리즘을 사용하였고, 국소최적해를 만났을 때 이를 벗어난 다른 연결가중치의 설정을 위해 동적터널링 시스템에 기초한 후향전파 알고리즘을 조합한 학습 알고리즘을 적용하였다. 제안된 방법을 패리티 검사 및 패턴분류 문제에 각각 적용하여 기존의 기울기 하강법에 기초한 후향전파 알고리즘 및 기울기 하강법과 동적터널링 시스템을 조합한 후향전파 알고리즘방법의 결과와 비교 고찰하여 제안된 방법이 다른 방법들 보다 학습성능에서 우수함을 나타내었다.

  • PDF

기울기하강과 동적터널링에 기반을 둔 학습알고리즘의 신경망을 이용한 영상데이터의 주요특징추출 (Principal Feature Extraction on Image Data Using Neural Networks of Learning Algorithm Based on Steepest Descent and Dynamic tunneling)

  • 조용현
    • 한국정보처리학회논문지
    • /
    • 제6권5호
    • /
    • pp.1393-1402
    • /
    • 1999
  • 본 논문에서는 새로운 학습알고리즘의 3층 전향 신경망을 이용한 입력데이터의 주요 특징추출에 대해서 제안하였다. 제안된 학습알고리즘에서에서는 빠른 수렴속도의 최적화가 가능하도록 하기 위하여 기울기하강의 역전파 알고리즘을 이용하고, 국소최적해를 만났을 때 이를 벗어난 새로운 연결가중치의 설정을 위하여 동적터널링의 역전파 알고리즘을 이용함으로써 빠른 수렴속도로 전역최적해로에 수렴되도록 학습시킬 수 있다. 제안된 학습 알고리즘을 이용한 다층신경망을 $12{\times}12$ 픽셀의 영상 데이터들과 $128{\times}128$ 픽셀의 Lenna 영상데이터를 대상으로 시뮬레이션한 결과, 단층신경망을 이용하는 Sanger 방법이나 측면연결을 가지는 단충신경망을 이용하는 Foldiak 방법 및 기울기하강에 기초를 둔 기존의 역전파 알고리즘을 이용한 다층신경망에 의한 결과와 비교할 때 더욱 우수한 수렴성능과 추출성능이 있음을 확인할 수 있었다.

  • PDF

Dynamic response characteristics of crossing tunnels under heavy-haul train loads

  • Dong, Jie;Zhong, Shuai;Wang, Hai-long;Wu, Zhi-hui
    • Geomechanics and Engineering
    • /
    • 제20권2호
    • /
    • pp.103-112
    • /
    • 2020
  • The dynamic response of crossing tunnels under heavy-haul train loads is still not fully understood. In this study, based on the case of a high-speed tunnel underneath an existing heavy-haul railway tunnel, a model experiment was performed to research the dynamic response characteristics of crossing tunnels. It is found that the under-crossing changes the dynamic response of the existing tunnel and surrounding rock. The acceleration response of the existing tunnel enhances, and the dynamic stress of rock mass between crossing tunnels decreases after the excavation. Both tunneling and the excitation of heavy-haul train loads stretch the tunnel base, and the maximum tensile strain is 18.35 µε in this model test. Then, the measured results were validated by numerical simulation. Also, a parametric study was performed to discuss the influence of the relative position between crossing tunnels and the advanced support on the dynamic behavior of the existing tunnel, where an amplifying coefficient of tunnel vibration was introduced to describe the change in acceleration due to tunneling. These results reveal the dynamic amplifying phenomenon of the existing tunnel during the new tunnel construction, which can be referred in the dynamic design of crossing tunnels.

이동 서비스를 위한 터널링 기법의 구현과 성능분석 (Implementation and Analysis of Tunneling Method for Mobile Service)

  • 천정훈;정진우강현국
    • 대한전자공학회:학술대회논문집
    • /
    • 대한전자공학회 1998년도 추계종합학술대회 논문집
    • /
    • pp.47-50
    • /
    • 1998
  • In recent years, it is required that computing support mobile user with computer. The advantage of mobile computing is that users may access all their applications from any location, whether they are in another building or a different state. So, Internet combines with mobile computing technology to make new communication environment for supporting mobility. The research for solving the problem of mobility is actively in progress. This paper describes the implementation of tunneling method for flexible bypass between specific region. Tunneling method provide mobile service to mobile hosts. IP datagram's address tranform method is IP-within-IP encapsulation by which an IP datagram may be encapsulated within an IP datagram. The developed IP-within-IP protocol can provide not only enhanced performance because it is implemented in kernel mode, but also convenience of usage to the application developers because it gives user interface as a dynamic link library. Verification of IP packet tunneling was text file transfer program.

  • PDF

Dynamic evolution characteristics of water inrush during tunneling through fault fracture zone

  • Jian-hua Wang;Xing Wan;Cong Mou;Jian-wen Ding
    • Geomechanics and Engineering
    • /
    • 제37권2호
    • /
    • pp.179-187
    • /
    • 2024
  • In this paper, a unified time-dependent constitutive model of Darcy flow and non-Darcy flow is proposed. The influencing factors of flow velocity are discussed, which demonstrates that permeability coefficient is the most significant factor. Based on this, the dynamic evolution characteristics of water inrush during tunneling through fault fracture zone is analyzed under the constant permeability coefficient condition (CPCC). It indicates that the curves of flow velocity and hydrostatic pressure can be divided into typical three stages: approximate high-velocity zone inside the fault fracture zone, velocity-rising zone near the tunnel excavation face and attenuation-low velocity zone in the tunnel. Furthermore, given the variation of permeability coefficient of the fault fracture zone with depth and time, the dynamic evolution of water flow in the fault fracture zone under the variable permeability coefficient condition (VPCC) is also studied. The results show that the time-related factor (α) affects the dynamic evolution distribution of flow velocity with time, the depth-related factor (A) is the key factor to the dynamic evolution of hydrostatic pressure.

Switching and sensing molecular spins by chemical reactions on metal surfaces

  • Kahng, Se-Jong
    • 한국진공학회:학술대회논문집
    • /
    • 한국진공학회 2015년도 제49회 하계 정기학술대회 초록집
    • /
    • pp.63.2-63.2
    • /
    • 2015
  • Controlling and sensing spin states of magnetic molecules such as metallo-porphyrins at the single molecule level is essential for spintronic molecular device applications. Axial coordinations of diatomic molecules to metallo-porphyrins also play key roles in dynamic processes of biological functions such as blood pressure control and immune response. However, probing such reactions at the single molecule level to understand their physical mechanisms has been rarely performed. Here we present on our single molecule association and dissociation experiments between diatomic and metallo-porphyrin molecules on Au(111) describing its adsorption structures, spin states, and dissociation mechanisms. We observed bright ring shapes in NO adsorbed metallo-porphyrin compelxes and explained them by considering tilted binding and precession motion of NO. Before NO exposure, Co-porphryin showed a clear zero-bias peak in scanning tunneling spectroscopy, a signature of Kondo effect in STS, whereas after NO exposures it formed a molecular complex, NO-Co-porphyrin, that did not show any zero-bias feature implying that the Kondo effect was switched off by binding of NO. Under tunneling junctions of scanning tunneling microscope, both positive and negative energy pulses. From the observed power law relations between dissociation rate and tunneling current, we argue that the dissociations were inelastically induced with molecular orbital resonances. Our study shows that single molecule association and dissociation can be used to probe spin states and reaction mechanisms in a variety of axial coordination between small molecules and metallo-porphyrins.

  • PDF