• Structural Engineering and Mechanics
• /
• 제24권2호
• /
• pp.213-225
• /
• 2006

The propagation of non-uniformly modulated, evolutionary random waves in viscoelastic, transversely isotropic, stratified materials is investigated. The theory is developed in the context of a multi-layered soil medium overlying bedrock, where the material properties of the bedrock are considered to be much stiffer than those of the soil and the power spectral density of the random excitation is assumed to be known at the bedrock. The governing differential equations are first derived in the frequency/wave-number domain so that the displacement response of the ground may be computed. The eigen-solution expansion method is then used to solve for the responses of the layers. This utilizes the precise integration method, in combination with the extended Wittrick-Williams algorithm, to obtain all the eigen-solutions of the ordinary differential equation. The recently developed pseudo-excitation method for structural random vibration is then used to determine the solution of the layered soil responses.

• 대한전기학회논문지:전기기기및에너지변환시스템부문B
• /
• 제49권8호
• /
• pp.557-564
• /
• 2000

A very simple control approach using neural network for the robust position control of a Permanent Magnet Synchronous Motor(PMSM) is presented. The linear quadratic controller plus feedforward neural network is employed to obtain the robust PMSM system approximately linearized using field-orientation method for an AC servo. The neural network is trained in on-line phases and this neural network is composed by a feedforward recall and error back-propagation training. Since the total number of nodes are only eight, this system can be easily realized by the general microprocessor. During the normal operation, the input-output response is sampled and the weighting value is trained multi-times by error back-propagation method at each sample period to accommodate the possible variations in the parameters or load torque. In addition, the robustness is also obtained without affecting overall system response. This method is realized by a floating-point Digital Signal Processor DS1102 Board (TMS320C31).

• Steel and Composite Structures
• /
• 제31권4호
• /
• pp.341-359
• /
• 2019

Non-monolithic concrete structural connections are commonly used both in new constructions and retrofitted structures where anchors are used for connections. Often, flaws are present in anchor system due to poor workmanship and deterioration; and methods available to check the quality of the composite system afterward are very limited. In case of presence of flaw, load transfer mechanism inside the anchor system is severely disturbed, and the load carrying capacity drops drastically. This raises the question of safety of the entire structural system. The present study proposes a wave propagation technique to assess the integrity of the anchor system. A chemical anchor (embedded in concrete) composite system comprising of three materials viz., steel (anchor), polymer (adhesive) and concrete (base) is considered for carrying out the wave propagation studies. Piezoelectric transducers (PZTs) affixed to the anchor head is used for actuation and the PZTs affixed to the surrounding concrete surface of the concrete-anchor system are used for sensing the propagated wave through the anchor interface to concrete. Experimentally validated finite element model is used to investigate three types of composite chemical anchor systems. Studies on the influence of geometry, material properties of the medium and their distribution, and the flaw types on the wave signals are carried out. Temporal energy of through time domain differentiation is found as a promising technique for identifying the flaws in the multi-layered composite system. The present study shows a unique procedure for monitoring of inaccessible but crucial locations of structures by using wave signals without baseline information.

• 한국강구조학회 논문집
• /
• 제16권4호통권71호
• /
• pp.443-452
• /
• 2004

인간의 뇌와 유사한 병렬 연산 모델을 활용하여 다양하고 복잡한 비선형적인 문제에 효과적으로 연관관계를 조직화 할 수 있는 인공신경망에 관한 연구가 근래에 공학의 넓은 분야에서 도입되고 그에 따른 많은 성과가 나타나고 있다. 본 연구에서는 설계자의 판단력과 경험에 의존 하던 기존의 예비구조설계 단계에 효과적인 인공신경망을 적용하여 예비 구조설계 단계에 컴퓨터를 이용한 정형화된 방법을 제시하고자 한다. 이를 위해 각 구조물의 일반적인 설계과정에 따른 다단계 신경망을 제시하고 인공신경망의 학습은 역전파알고리즘과 유전알고리즘을 적용하여 예비구조설계의 원형을 구현한다. 이와 같이 구성된 다단계 신경망을 사장교의 예비구조설계 단계에 활용하여 본 연구의 적용성과 두가지 학습기법에 따른 결과를 비교 분석 한다.

• 디지털융복합연구
• /
• 제18권7호
• /
• pp.31-36
• /
• 2020

최근 들어, 대규모의 피해를 동반하는 각종 재난, 사고 등이 증가하고 있다. 재난은 불확실성을 지닌 사고로 국민의 생명과 안전, 재산보호 등에 직접적인 영향을 미친다. 따라서, 다양한 재난에 대한 예방, 대응, 복구 등의 안전관리 체계를 구축, 운영하는 것이 필요하다. 따라서, 본 논문에서는 재난 사고에 대한 체계적인 대응을 위해 스마트 환경에서의 실시간 재난 안전 관리 솔루션을 설계하였다. 이를 위해, 스마트 디바이스를 통해 상황실과 유관기관, 전문가 등에게 1:1 또는 1:N 상황 전파를 수행하여 다자간 정보 공유 및 소통으로 신속하고 적절하게 대응할 수 있도록 솔루션을 구성하였다. 즉, 실시간·다자간 HD 영상 송출, 모바일 형식의 보고서 관리, 음성·문자 상황 전파, 위치정보 공유, 녹화 및 이력관리, 보안 등의 기능을 적용한 솔루션을 설계하였다.

• 한국지능시스템학회:학술대회논문집
• /
• 한국퍼지및지능시스템학회 2000년도 춘계학술대회 학술발표 논문집
• /
• pp.47-50
• /
• 2000

In this paper, an optimal identification method using Multi-FNN(Fuzzy-Neural Network) is proposed for model ins of nonlinear complex system. In order to control of nonlinear process with complexity and uncertainty of data, proposed model use a HCM clustering algorithm which carry out the input-output data preprocessing function and Genetic Algorithm which carry out optimization of model. The proposed Multi-FNN is based on Yamakawa's FNN and it uses simplified inference as fuzzy inference method and Error Back Propagation Algorithm as learning rules. HCM clustering method which carry out the data preprocessing function for system modeling, is utilized to determine the structure of Multi-FNN by means of the divisions of input-output space. Also, the parameters of Multi-FNN model such as apexes of membership function, learning rates and momentum coefficients are adjusted using genetic algorithms. Also, a performance index with a weighting factor is presented to achieve a sound balance between approximation and generalization abilities of the model, To evaluate the performance of the proposed model, we use the time series data for gas furnace and the numerical data of nonlinear function.

• Ocean and Polar Research
• /
• 제43권4호
• /
• pp.295-306
• /
• 2021

A field scale productivity analysis is required for the development of gas hydrate in marine sedimentary layers to verify the field applicability of production techniques and to improve productivity. In this study, the productivity resulting from the application of depressurization using multi-wells for the development of gas hydrate in the Ulleung Basin, East Sea of Korea, was determined. A numerical analysis model reflecting the conditions of candidate sites for the Ulleung Basin was constructed, and the productivity and dissociation behavior were comparatively analyzed. The pressure propagation and gas hydrate dissociation region by the multi-wells were wider and the productivity was higher than that of a single well. Different depressurization effects according to the spacing of multi-wells affected productivity. The results provide basic data for productivity analysis when establishing a field test production plan for the Ulleung Basin.

• 한국정보처리학회:학술대회논문집
• /
• 한국정보처리학회 2022년도 추계학술발표대회
• /
• pp.308-310
• /
• 2022

With the rise of the Al wave, task-oriented dialogue systems have become one of the popular research directions in academia and industry. Currently, task-oriented dialogue systems mainly adopt pipelined form, which mainly includes natural language understanding, dialogue state decision making, dialogue state tracking and natural language generation. However, pipelining is prone to error propagation, so many task-oriented dialogue systems in the market are only for single-round dialogues. Usually single- domain dialogues have relatively accurate semantic understanding, while they tend to perform poorly on multi-domain, multi-round dialogue datasets. To solve these issues, we developed a paddle-based multi-domain task-oriented Chinese dialogue system. It is based on NEZHA-base pre-training model and CrossWOZ dataset, and uses intention recognition module, dichotomous slot recognition module and NER recognition module to do DST and generate replies based on rules. Experiments show that the dialogue system not only makes good use of the context, but also effectively addresses long-term dependencies. In our approach, the DST of dialogue tracking state is improved, and our DST can identify multiple slotted key-value pairs involved in the discourse, which eliminates the need for manual tagging and thus greatly saves manpower.

• 한국통신학회논문지
• /
• 제28권10A호
• /
• pp.807-815
• /
• 2003

MultiCode-CDMA(MC-CDMA) 시스템은 채널수가 증가할수록 신호의 PAPR(Peak to Average Power Ratio)이 증가하게 되어 증폭기의 비선형 특성에 의해 시스템 성능이 열화된다. MultiPhase Clipped MultiCode CDMA(MP-CDMA)[1] 시스템은 이러한 증폭기의 비선형 특성에 의한 시스템 성능 열화를 상쇄한다. 본 논문에서는 변조방식을 MPSK가 아닌 차동 부호화된 MPSK 방식을 사용했고 이렇게 변형된 시스템을 DMP-CDMA 시스템이라고 부를것이다. 차동 검출 방식을 적용한 DMP-CDMA 시스템은 위상 보정을 위한 모듈이 필요없고, 파일럿(pilot) 신호를 위한 추가적인 데이터 전송이 필요 없다는 장점을 가진다. 그러나 동기 검출을 사용한 시스템에 비해서 약 4.0dB의 성능열화를 보인다. 차동 검출 방식을 사용한 DMP-CDMA 시스템의 성능 열화를 보상하기 위해서 다중 심볼 차동 검출 방식을 적용한 결과 차동 검출 방식에 비해서 약 3.6dH의 성능 개선 효과를 보았다.

• 한국연소학회:학술대회논문집
• /
• 한국연소학회 제26회 KOSCO SYMPOSIUM 논문집
• /
• pp.85-91
• /
• 2003

The lift-off characteristics of the triple flame within a diverging duct have been studied experimentally using a multi-slot burner, which can control the concentration gradient and the mean velocity independently. In this experiment the triple flame was stabilized successfully in lift-off condition and flame stabilization with a duct or without a duct, lift-off heights, and some other characteristics were examined for propane flame. It was examined that the effects with various concentration gradient and mean velocity on the triple flame. It was found that minimum value of the lift-off heights exist at a certain concentration gradient for constant mean velocity and flame with a duct is more stable than that without. Moreover the propagation velocity of the flame becomes maximum at a certain concentration gradient regardless of mean velocity.