• 제목/요약/키워드: simulated network

검색결과 938건 처리시간 0.029초

신경망 학습의 일반화 성능향상을 위한 인자들의 결합효과 (The Joint Effect of factors on Generalization Performance of Neural Network Learning Procedure)

  • 윤여창
    • 정보처리학회논문지B
    • /
    • 제12B권3호
    • /
    • pp.343-348
    • /
    • 2005
  • 본 연구에서는 신경망 학습의 일반화 성능과 학습속도를 개선시키기 위한 인자들의 결합 효과를 살펴본다. 신경망 학습에서 중요한 평가 척도로서 여기서 고려하는 인자들에는 초기 가중값의 범위와 학습률 그리고 계수조정 등이 있다. 특히 초기 가중값과 학습률을 고정시킨 후 새롭게 조정된 계수들을 단계적으로 변화시키는 새로운 인자 결합방법을 이용한다. 이를 통하여 신경망 학습량과 학습속도를 비교해 보고, 계수조정을 통한 개선된 학습 영향을 살펴본다. 그리고 비선형의 단순한 예제를 이용한 실증분석을 통하여 신경망 모형의 일반화 성능과 학습 속도 개선을 위한 각 인자들의 개별 효과와 결합 효과를 살펴보고 그 개선 방안을 논의한다.

Spatiotemporal Routing Analysis for Emergency Response in Indoor Space

  • Lee, Jiyeong;Kwan, Mei-Po
    • 한국측량학회지
    • /
    • 제32권6호
    • /
    • pp.637-650
    • /
    • 2014
  • Geospatial research on emergency response in multi-level micro-spatial environments (e.g., multi-story buildings) that aims at understanding and analyzing human movements at the micro level has increased considerably since 9/11. Past research has shown that reducing the time rescuers needed to reach a disaster site within a building (e.g., a particular room) can have a significant impact on evacuation and rescue outcomes in this kind of disaster situations. With the purpose developing emergency response systems that are capable of using complex real-time geospatial information to generate fast-changing scenarios, this study develops a Spatiotemporal Optimal Route Algorithm (SORA) for guiding rescuers to move quickly from various entrances of a building to the disaster site (room) within the building. It identifies the optimal route and building evacuation bottlenecks within the network in real-time emergency situations. It is integrated with a Ubiquitous Sensor Network (USN) based tracking system in order to monitor dynamic geospatial entities, including the dynamic capacities and flow rates of hallways per time period. Because of the limited scope of this study, the simulated data were used to implement the SORA and evaluate its effectiveness for performing 3D topological analysis. The study shows that capabilities to take into account detailed dynamic geospatial data about emergency situations, including changes in evacuation status over time, are essential for emergency response systems.

차량정보 분석과 제스처 인식을 위한 AVN 소프트웨어 구현 (Development of AVN Software Using Vehicle Information for Hand Gesture)

  • 오규태;박인혜;이상엽;고재진
    • 한국통신학회논문지
    • /
    • 제42권4호
    • /
    • pp.892-898
    • /
    • 2017
  • 본 논문은 차량 내 AVN(Audio Video Navigation)에서 차량정보 분석과 제스처 인식이 가능한 소프트웨어 구조를 설계하고 구현 방법을 서술한다. 설계된 소프트웨어는 차량정보 분석을 위해 CAN(Controller Area Network) 통신 데이터 분석 모듈을 구현하여 차량의 주행 상태를 분석했다. AVN 소프트웨어는 분석된 정보를 웨어러블 디바이스의 제스처 정보와 융합토록 했다. 도출된 융합정보는 운전자의 명령 수행 단계로 매칭하고 서비스를 지원하는데 사용됐다. 설계된 AVN 소프트웨어는 기성 제품과 유사한 환경의 HW 플랫폼 상에 구현되어 차량 주행 상황과 동일하게 모사된 상황에서의 차량정보분석, 제스처 인식 수행 등의 기능을 지원함을 확인했다.

신경회로망을 이용한 GIS용 단로기의 이상신호 패턴분석 (Fatten Analysis of Trouble Signal on DS for GIS using Neural Network)

  • 김종서;이은석;천종철
    • 한국전기전자재료학회논문지
    • /
    • 제16권12S호
    • /
    • pp.1310-1315
    • /
    • 2003
  • GIS(Gas Insulated Switchgear) is power equipment with excellent dielectric strength and is economy merit in high confidence and stability. But, because of structural characteristics, this is difficult to monitor externally and provide much loss in the event of an accident. Recently, because equipment of GIS is occurring problem of confidence used for a long time, development of diagnosis technique have been importantly recognized. therefore, measurement and analysis of PD much has been generally used for equipment of GIS. But, in case of measurement of PD at field, real trouble signals is difficult to classify noise. Accordingly, a variety of trouble conditions for DS are simulated, and detected signals are analyzed by the application of electrical and mechanical methods. For this analysis, detected signals are accumulated according to phase-magnitude with the application of induction sensor, and then we analyzed the characteristics. For the simulation experiment, we make DS for 170KV GIS and analyze the characteristics of detected signals with the application of neural network algorithm. last, we have measured DS for GIS at field, and then have analyzed detected signals.

Solution Structure of the Cytoplasmic Domain of Syndecan-3 by Two-dimensional NMR Spectroscopy

  • Yeo, In-Young;Koo, Bon-Kyung;Oh, Eok-Soo;Han, Inn-Oc;Lee, Weon-Tae
    • Bulletin of the Korean Chemical Society
    • /
    • 제29권5호
    • /
    • pp.1013-1017
    • /
    • 2008
  • Syndecan-3 is a cell-surface heparan sulfate proteoglycan, which performs a variety of functions during cell adhension process. It is also a coreceptor for growth factor, mediating cell-cell and cell-matrix interaction. Syndecan-3 contains a cytoplasmic domain potentially associated with the cytoskeleton. Syndecan-3 is specifically expressed in neuron cell and has related to neuron cell differentiation and development of actin filament in cell migration. Syndecans each have a unique, central, and variable (V) region in their cytoplasmic domains. And that region of syndecan-3 may modulate the interactions of the conserved C1 regions of the cytoplasmic domains by tyrosine phosphorylation. Cytoplasmic domain of syndecan-3 has been synthesized for NMR structural studies. The solution structure of syndecan-3 cytoplasmic domain has been determined by two-dimensional NMR spectroscopy and simulated-annealing calculation. The cytoplasmic domain of the syndecan proteins has a tendency to form a dimmer conformation with a central cavity, however, that of syndecan-3 demonstrated a monomer conformation with a flexible region near C-terminus. The structural information might add knowledge about the structure-function relationships among syndecan proteins.

이동 백홀 네트워크에서 QoS 기능 (QoS Functions in Mobile Backhaul Network)

  • 박천관
    • 한국인터넷방송통신학회논문지
    • /
    • 제13권5호
    • /
    • pp.101-105
    • /
    • 2013
  • 본 논문은 셀 사이트에서 다양한 트래픽을 수용하기 위한 이동 백홀 네트워크에서 QoS 기능에 관한 것이다. RAN 시스템의 스위칭 기능에 할당된, 이더넷 프레임, IP 패킷, 그리고 ATM 셀과 같은 트래픽은 세그먼트된 후 이동백홀 네트워크로 전달되기 전에 캡슐화 된다. ISP는 표준의 pseudowire 캡슐화를 통하여 all IP RAN 상에 음성 및 HSPA와 같은 모든 세대의 트래픽을 통합할 수 있다. 이것들은 다양한 QoS 방법뿐만 아니라 포괄적인 모니터링 및 진단 능력을 통하여 강화될 수 있다. 따라서 본 논문에서는 이와 같은 동작상에서 QoS 기능들이 서로 다른 갭슐 방법에 따라 시뮬레이션 되었다.

An On-Chip Differential Inductor and Its Use to RF VCO for 2 GHz Applications

  • Cho, Je-Kwang;Nah, Kyung-Suc;Park, Byeong-Ha
    • JSTS:Journal of Semiconductor Technology and Science
    • /
    • 제4권2호
    • /
    • pp.83-87
    • /
    • 2004
  • Phase noise performance and current consumption of Radio Frequency (RF) Voltage-Controlled Oscillator (VCO) are largely dependent on the Quality (Q) factor of inductor-capacitor (LC) tank. Because the Q-factor of LC tank is determined by on-chip spiral inductor, we designed, analyzed, and modeled on-chip differential inductor to enhance differential Q-factor, reduce current consumption and save silicon area. The simulated inductance is 3.3 nH and Q-factor is 15 at 2 GHz. Self-resonance frequency is as high as 13 GHz. To verify its use to RF applications, we designed 2 GHz differential LC VCO. The measurement result of phase noise is -112 dBc/Hz at an offset frequency of 100 kHz from a 2GHz carrier frequency. Tuning range is about 500 MHz (25%), and current consumption varies from 5mA to 8.4 mA using bias control technique. Implemented in $0.35-{\mu}m$ SiGe BiCMOS technology, the VCO occupies $400\;um{\times}800\;um$ of silicon area.

KMTNet 실시간 자료처리 파이프라인 개발 (DEVELOPMENT OF REAL-TIME DATA REDUCTION PIPELINE FOR KMTNet)

  • 김동진;이충욱;김승리;박병곤
    • 천문학논총
    • /
    • 제28권1호
    • /
    • pp.1-6
    • /
    • 2013
  • Real-time data reduction pipeline for the Korea Microlensing Telescope Network (KMTNet) was developed by Korea Astronomy and Space Science Institute (KASI). The main goal of the data reduction pipeline is to find variable objects and to record their light variation from the large amount of observation data of about 200 GB per night per site. To achieve the goal we adopt three strategic implementations: precision pointing of telescope using the cross correlation correction for target fields, realtime data transferring using kernel-level file handling and high speed network, and segment data processing architecture using the Sun-Grid engine. We tested performance of the pipeline using simulated data which represent the similar circumstance to CTIO (Cerro Tololo Inter-American Observatory), and we have found that it takes about eight hours for whole processing of one-night data. Therefore we conclude that the pipeline works without problem in real-time if the network speed is high enough, e.g., as high as in CTIO.

On-line Monitoring and Control of Substrate Concentrations in Biological Processes by Flow Injection Analysis Systems

  • Rhee, Jong-Il;Adnan Ritzka;Thomas Scheper
    • Biotechnology and Bioprocess Engineering:BBE
    • /
    • 제9권3호
    • /
    • pp.156-165
    • /
    • 2004
  • Concentrations of substrates, glucose, and ammionia in biological processes have been on-line monitored by using glucose-flow injection (FIA) and ammonia-FIA systems. Based on the on-line monitored data the concentrations of substrates have been controlled by an on-off controller, a PID controller, and a neural network (NN) based controller. A simulation program has been developed to test the control quality of each controller and to estimate the control parameters. The on-off controller often produced high oscillations at the set point due to its low robustness. The control quality of a PID controller could have been improved by a high analysis frequency and by a short residence time of sample in a FIA system. A NN-based controller with 3 layers has been developed, and a 3(input)-2(hidden)-1(output) network structure has been found to be optimal for the NN-based controller. The performance of the three controllers has been tested in a simulated process as well as in a cultivation process of Saccharomyces cerevisiae, and the performance has also been compared to simulation results. The NN-based controller with the 3-2-1 network structure was robust and stable against some disturbances, such as a sudden injection of distilled water into a biological process.

가스분사 방식 LPG 엔진의 연료공급시스템 관로 유동해석 (Fluid Network Analysis for the Fuel-Supply Systems of Gaseous-Injection-Type LPG Engines)

  • 윤정의
    • 대한기계학회논문집B
    • /
    • 제35권10호
    • /
    • pp.1019-1024
    • /
    • 2011
  • LPG 분사시스템의 경우 액상 및 기상 인젝터를 사용하는 두 가지 방식이 있으며, 가스 분사 방식의 경우 내구성 및 가격 측면에서 장점이 있지만 가스의 압축성 특성으로 인한 정밀 유량제어에 어려움이 있다. 본 연구에서는 가스분사 방식 LPG 연료 분사 시스템에서 정밀 유량 제어를 위한 수단으로 헬름홀쯔 공진기를 사용하는 방안을 제시하였다. 그리고 상용 유동해석 프로그램인 Flowmaster를 사용하여 대상 자동차의 연료 시스템에 대해 유동해석을 통해 최적의 공진기를 설계방안을 설계하였으며, 공진기 설치 후 개선 효과를 제시하였다.