• Bulletin of the Society of Naval Architects of Korea
• /
• v.31 no.1
• /
• pp.39-42
• /
• 1994

인공 신경망은 빠른 속도와 안정성 등의 많은 장점을 갖고 있기 때문에 최근 들어서 여러 분야 에서 그 응용이 활발히 연구되고 있다. 인공 신경망의 한 모델인 홉필드 네트워크는 네트워크의 에너지를 최소화시키는 방향으로 네트워크의 상태를 바꾸며, 최소 에너지 상태에서 안정 상태를 유지하는 특징을 갖고 있다. 이러한 흡필드 네트워크의 특징은 흡필드 네트워크를 최적화 문 제에 적용시킬 수 있는 가능성을 제시하고 있다. 기존의 최적화 기법은 기본적으로 국부적인 탐색 기법을 사용하기 때문에, 전역적 최적해를 구하기 위해 초기점을 달리하여 여러번 계산을 수행하여 그 중 가장 좋은 결과를 취하는 방법을 사용하여야 한다. 따라서 이러한 방법은 초 기점의 선택이 결과에 큰 영향을 미치게 되는데, 설계 변수가 많고 제한 조건이 복잡할 경우 초기점 선택에 어려움이 따른다. 본 연구에서는 흡필드 네트워크와 시뮬레이티드 어닐링을 결 합하여 전역적 최적해를 찾는 기법으로서 뉴드-옵티마이저 모델을 제시하고 있다.

• Information and Communications Magazine
• /
• v.33 no.7
• /
• pp.18-24
• /
• 2016

본고에서는 MPEG에서 설계하여 ISO 표준화한 실시간/저지연 미디어 전송 기술인 MMT (MPEG Media Transport) 에 대해 알아본다. MMT는 인터넷 환경과 통합된 차세대 방송 시스템에의 적용을 목표로 설계된 기술로, IP 기반 전송 프로토콜의 사용과 HTML5 기반 서비스 제공을 가정하여 MPEG2-TS의 장점을 계승하면서 이종망 및 다기기 서비스의 지원을 고려하여 설계 되었다. 따라서 MMT는 고품질의 UHD TV서비스를 안정적으로 제공함과 동시에 인터넷 활용 부가 서비스 제공을 목적으로 하는 차세대 지상파 방송에 있어 최적화된 표준 기술이라 할 수 있다.

• Information and Communications Magazine
• /
• v.15 no.5
• /
• pp.88-98
• /
• 1998

지금까지 ATM 망의 생존도에 대하여 연구되어 온 관련 모형 및 알고리듬은 주로 가상 경로의 특징을 제대로 적용하지 않고, 기존의 STM 망 모형의 변형을 이용한 연구가 대부분이었다. 또한 장애가 일어났을 때 ATM의 OAM 셀을 사용해 장애를 어떻게 효율적으로 알리고 복구 확인을 하는 차원의 연구가 많은 반면, 복구가 망 전체 차원에서 어떻게 자원을 최적화로 사용하고 관리 및 계획의 객체를 정의할 지 등에 대한 체계적인 연구나 모형화가 시작되지 못해왔다. 이 논문에서는 ATM 망의 가상경로 개념을 사용한 생전도 확보 기법에 대하여 지금까지의 연구 동향 및 향후 연구방향에 대하여 기술한다.

• The KIPS Transactions:PartB
• /
• v.13B no.4 s.107
• /
• pp.429-438
• /
• 2006

Neural Networks(NNs) are applied for solving a wide variety of nonlinear problems in several areas, such as image processing, pattern recognition etc. Although NN can be simulated by using software, many potential NN applications required real-time processing. Thus they need to be implemented as hardware. The hardware implementation of multi-layer perceptrons(MLPs) in several kind of NNs usually uses a fixed-point arithmetic due to a simple logic operation and a shorter processing time compared to the floating-point arithmetic. However, the fixed-point arithmetic-based MLP has a drawback which is not able to apply the MLP software that use floating-point arithmetic. We propose a design method for MLPs which has the floating-point arithmetic-based fully-pipelining architecture. It has a processing speed that is proportional to the number of the hidden nodes. The number of input and output nodes of MLPs are generally constrained by given problems, but the number of hidden nodes can be optimized by user experiences. Thus our design method is using optimized number of hidden nodes in order to improve the processing speed, especially in field of a repeated processing such as image processing, pattern recognition, etc.

• Journal of Korean Society of Transportation
• /
• v.19 no.1
• /
• pp.89-99
• /
• 2001

The conventional studies on equilibrium network design problem(ENDP) with fixed travel demand models assume that the future OD travel demand might not be changed even if the structure and the capacity of the network are improved. But this fixed demand assumption may loose its validity in the long-range network design because OD travel demand actually shifts with the network service level. Thus, it is desirable to involve the variable travel demand which is determined endogenously in the model in the optimal network design. In this paper a hi-level model formulation and solution procedure for ENDP with variable travel demand are presented. Firstly It is considered how to measure the net user benefits to be derived from the improved in link capacities, and the equilibrium network design problem considered here is to maximize the increase of net user benefit which results from a set of lift capacity enhancements within the budget constraints, while the OD travel demands and link travel times are obtained by solving the lower level network equilibrium problem with variable demand. And secondly sensitivity analysis is carried out to find the links to which the network equilibrium flow pattern is the most sensitive. Finally numerical example with simple network is carried out to test the validity of the model.

• The Journal of Korean Institute of Communications and Information Sciences
• /
• v.23 no.9A
• /
• pp.2221-2229
• /
• 1998

Handoff is an important parameter for system performance and capacity analysis in CDMA cell planning. In this paper, simulation tool for analyzing the effects of soft handoff on cell planning of CDMA network is developed by studying propagation model in 1.9GHz frequency specrum, soft hadnoff algorithm given by IS-95, and several parameters for CDMA cell planning. By using the tool, soft handoff gain and effects of soft handoff on cDMA system performance and capacity are analyzed. Research results and simulation tool developed in this paper can be used for optimal CDMA cell planning for PCS and IMT-2000 in 1.9GHz frequency spectrum such as locating base station and parameter optimization.

• Journal of Korean Society of Steel Construction
• /
• v.16 no.6 s.73
• /
• pp.819-832
• /
• 2004

This study describes an efficient and facile method for configuration optimum design of structures. One of the ways to achieve numerical shape representation and the selection of design variables is using the design element concept. Using this technique, the number of design variables could be drastically reduced. Isoparametric mapping was utilized to automatically generate the finite element mesh during the optimization process, and this made it possible to easily calculate the derivatives of the coordinates of generated finite element nodes w.r.t. the design variables. For the structural analysis, finite element analysis was adopted in the optimization procedure, and two different techniques(the deterministic method, a modified method of feasible direction; and the stochastic method, a genetic algorithms) were applied to obtain the minimum volumes and section areas for an efficient configuration optimization procedure. Futhermore, spline interpolation was introduced to present a realistic optimum configuration that meet the manufacturing requirements. According to the results of several numerical examples(steel structures), the two techniques suggested in this study simplified the process of configuration optimum design of structures, and yielded improved objective function values with a robust convergence rate. This study's applicability and capability have therefore been demonstrated.

• Journal of the Korean Society for Aeronautical & Space Sciences
• /
• v.47 no.3
• /
• pp.177-186
• /
• 2019

Wind power tower has been used to augment the performance of VAWT (Vertical Axis Wind Turbine). However, inappropriately designed wind power tower could reduce the performance of VAWT. Hence, an optimization study was conducted on a wind power tower. Six design variables were selected, such as the outer radius and the inner radius of the guide wall, the adoption of the splitter, the inner radius of the splitter, the number of the guide wall and the circumferential angle. For the objective function, the periodic averaged torque obtained at the VAWT was selected. In the optimization, Design of Experiment (DOE), Genetic Algorithm (GA), and Artificial Neural Network (ANN) have been applied in order to avoid a localized optimized result. The ANN has been continuously improved after finishing the optimization process at each generation. The performance of the VAWT was improved more than twice when it operated within the optimized wind power tower compared to that obtained at a standalone.

• Journal of Advanced Navigation Technology
• /
• v.19 no.1
• /
• pp.53-59
• /
• 2015

We consider the problems that consist of designing time, establishment cost, delay time and reliability in designing a mesh network when given link costs and traffic requirements between nodes. Designing time, establishment cost and delay time are less, reliability is higher in designing a mesh network. One of the problems designing time is solved by mesh network topology optimization and routing (MENTOR) algorithm that Aaron Kershenbaum propose, but the others remain. In this paper we propose a new mesh network design algorithm with small computational complexity that the others are solved. The result of the proposed algorithm is better than MENTOR's in total establishment cost, delay time and reliability.

• Journal of Korea Water Resources Association
• /
• v.31 no.6
• /
• pp.685-694
• /
• 1998

Protection of groundwater resources from contamination has been of increasing concern throughout the past decades. In practice, however, groundwater monitoring is performed based on the experience and intuition of experts or on the convenience. In dealing with groundwater contamination, we need to know what contaminants have the potential to threat the water quality and the distribution and concentration of the plumes. Monitoring of the subsurface environment through remote geophysical techniques or direct sampling from wells can provide such information. Once known, the plume can be properly managed. Evaluation of existing methodologies for groundwater monitoring network design revealed that one should select an appropriate design method based on the purpose of the network and the availability of field information. Integer programming approach, one of the general purpose network design tools, and a cost to-go function evaluation approach for special purpose network design were tested for field applicability. For the sam contaminated aquifer, two approaches resulted in different well locations. The amount of information, however, was about the same.