• 한국시스템다이내믹스연구
• /
• 제16권2호
• /
• pp.33-52
• /
• 2015

This study aims to comprehensively understand the dynamics of the community network of Community Child Centers and further find out the measures to activate its community network based on the Systems Thinking. The contents of the study are as follows. Firstly, it examines the existing studies on the community network of Community Child Centers and presents the major variables to understand the situation of its community network. Secondly, it analyzes the structure of its causation in order to understand the dynamics of its community network. Lastly, it concludes with the suggestions to activate its community network based on its feedback structure presented in the causal loop diagrams. This study is expected to make a useful and basic material as the first research to dynamically understand the community network issue of the Community Child Centers.

• E2M - 전기 전자와 첨단 소재
• /
• 제9권7호
• /
• pp.702-707
• /
• 1996

In this paper, in order to improve the withstand voltage properties of epoxy resin, IPN(interpenetrating polymer network) method was introduced and the influence was investigated. The single network structure specimen(E series), simultaneous interpenetrating polymer network specimen(EM series) and pseudo interpenetrating polymer network(EMP series) specimen were manufactured. In order to understand the internal structure properties, scanning electron microscopy method was utilized, and glass transition temperature was measured. Also, AC voltage dielectric breakdown strength, tensile strength and impact strength were measured to investigate the influence upon electrical and mechanical properties. As a result, it was confirmed that simultaneous interpenetrating polymer network specimen was the most execellent.

• Journal of Information Processing Systems
• /
• 제17권2호
• /
• pp.411-425
• /
• 2021

In this paper, a vehicle recognition algorithm based on deep convolutional neural network and compression dictionary is proposed. Firstly, the network structure of fine vehicle recognition based on convolutional neural network is introduced. Then, a vehicle recognition system based on multi-scale pyramid convolutional neural network is constructed. The contribution of different networks to the recognition results is adjusted by the adaptive fusion method that adjusts the network according to the recognition accuracy of a single network. The proportion of output in the network output of the entire multiscale network. Then, the compressed dictionary learning and the data dimension reduction are carried out using the effective block structure method combined with very sparse random projection matrix, which solves the computational complexity caused by high-dimensional features and shortens the dictionary learning time. Finally, the sparse representation classification method is used to realize vehicle type recognition. The experimental results show that the detection effect of the proposed algorithm is stable in sunny, cloudy and rainy weather, and it has strong adaptability to typical application scenarios such as occlusion and blurring, with an average recognition rate of more than 95%.

• Smart Structures and Systems
• /
• 제11권3호
• /
• pp.315-329
• /
• 2013

The construction of an experimental nonlinear structural model with little cost and unlimited repeatability for vibration control study represents a challenging task, especially for material nonlinearity. This paper reports the design, analysis and vibration control of a nonlinear structural experiment platform with adjustable hinges. In our approach, magnetorheological rotary brakes are substituted for the joints of a frame structure to simulate the nonlinear material behaviors of plastic hinges. For vibration control, a separate magnetorheological damper was employed to provide semi-active damping force to the nonlinear structure. A dynamic neural network was designed as a state observer to enable the feedback based semi-active vibration control. Based on the dynamic neural network observer, an adaptive fuzzy sliding mode based output control was developed for the magnetorheological damper to suppress the vibrations of the structure. The performance of the intelligent control algorithm was studied by subjecting the structure to shake table experiments. Experimental results show that the magnetorheological rotary brake can simulate the nonlinearity of the structural model with good repeatability. Moreover, different nonlinear behaviors can be achieved by controlling the input voltage of magnetorheological rotary damper. Different levels of nonlinearity in the vibration response of the structure can be achieved with the above adaptive fuzzy sliding mode control algorithm using a dynamic neural network observer.

• 한국구조물진단유지관리공학회 논문집
• /
• 제6권4호
• /
• pp.169-179
• /
• 2002

This paper presents an efficient algorithm for the estimation of damage location and severity in structure using Probabilistic Neural Network (PNN). Artificial neural network has been being used for damage assessment by many researchers, but there are still some barriers that must be overcome to improve its accuracy and efficiency. The major problems with the conventional neural network are the necessity of many training data for neural network learning and ambiguity in the relation of neural network architecture with convergence of solution. In this paper, PNN is used as a pattern classifier to overcome those problems in the conventional neural network. The basic idea of damage assessment algorithm proposed in this paper is that modal characteristics from a damaged structure are compared with the training patterns which represent the damage in specific element to determine how close it is to training patterns in terms of the probability from PNN. The training pattern that gives a maximum probability implies that the element used in producing the training pattern is considered as a damaged one. The proposed damage assessment algorithm using PNN is applied to a 2-span continuous beam model structure to verify the algorithm.

• 한국정보통신학회:학술대회논문집
• /
• 한국정보통신학회 2016년도 추계학술대회
• /
• pp.907-909
• /
• 2016

본 논문에서는 DTN에서 UAV 편대를 이용한 효율적인 네트워크 구조를 제안한다. 기존 MANET에서 불안정한 종단간 연결성으로 인한 문제를 해결하기 위해 제안된 DTN은 Store-Carry-Forward 방식을 사용하여 통신한다. Store-Carry-Forward 방식의 통신은 종단간의 연결 없이도 중계 노드가 확보되면 번들 계층을 통해 메시지를 저장하여 목적 노드까지 손실 없이 메시지를 전달한다. 이러한 라우팅 구조는 재해재난 상황이나 전쟁 지역과 같이 네트워크 기반시설이 없는 곳에서도 네트워크를 구성하는 데 유리하기 때문에 현재 많은 연구가 진행되고 있다. 본 논문은 그런 연구의 일환으로 UAV 편대를 이용하여 유동적인 환경에서 더 효율적인 네트워크 구조를 제안한다. 제안하는 방안은 센서로써 작동하는 이질적인 노드들의 환경 정보 이용해 UAV 분대를 조직하여 안정적인 통신이 가능하도록 한다.

• 한국방재학회:학술대회논문집
• /
• 한국방재학회 2007년도 정기총회 및 학술발표대회
• /
• pp.429-432
• /
• 2007

For smart structure technologies, the interests in wireless sensor networks for structural health monitoring are growing. The wireless sensor networks reduce the installation of the wire embedded in the whole structure and save the costs. But the wireless sensor networks have lots of limits and there are lots of researches and developments of wireless sensor and the network for data process. Most of the researches of wireless sensor network is applying to the civil engineering structure and the researches for the highrise building are required. And strain-based SHM gives the local damage information of the structures which acceleration-based SHM can not. In this paper, concept of wireless sensor network for structural health monitoring of highrise building is suggested. And verifying the feasibility of the strain-based SHM a strain sensor board has developed and tested by experiments.

• 융합신호처리학회논문지
• /
• 제2권2호
• /
• pp.111-118
• /
• 2001

본 연구에서는 패턴 인식용 다층 퍼셉트론 신경망을 순수 디지털 논리회로 모델로 구현할 수 있도록 새로운 논리뉴런의 구조, 디지털 정형 다층논리신경망 구조, 그리고 패턴인식의 응용을 위한 다단 다층논리 신경망 구조를 제안하고, 또한 제안된 구조는 매우 단순하면서도 효과적인 증가적인 가법적(Incremental Additive) 학습알고리즘이 존재함을 보였다.

• 전자공학회논문지B
• /
• 제28B권5호
• /
• pp.369-381
• /
• 1991

In this paper, design algorithms for SMP(Single Mated Pair) and MMP (Multipli Mated Pair) structure of CCS (Common Channel Signaling) network are proposed through the study of the structure of CCS network. High reliability and fast messagy transfer time are the most important requirements for the CCS network. Based on it, three parameters such as monotraffic, reliability (maximum isolated SP(Signalling Point) number when any two STP(Signalling Transfer Points) fail and total network cost are defined. And the proposed algorithms different from preexisted algorithm that minimizes total network cost, maximize monotraffic with two constraints, reliability and total network cost. Comparing the experimental results of the proposed algorithms with those of the preexisted algorithm that minimizes total network cost, shows that the proposed algorithms produce a more reliable topology that has more monotraffic and a little higher total network cost. Additionaly, with the results of the proposed algorithms, SMP and MMP structures are compared.

• 한국산학기술학회:학술대회논문집
• /
• 한국산학기술학회 2010년도 추계학술발표논문집 1부
• /
• pp.344-348
• /
• 2010

The intent of this paper is to describe a neural network structure called multi dynamic neural network(MDNN), and examine how it can be used in developing a learning scheme for computing robot inverse kinematic transformations. The architecture and learning algorithm of the proposed dynamic neural network structure, the MDNN, are described. Computer simulations are demonstrate the effectiveness of the proposed learning using the MDNN.