• 제목/요약/키워드: Location and scale parameters

검색결과 133건 처리시간 0.029초

결합된 파라메트릭 활성함수를 이용한 합성곱 신경망의 성능 향상 (Performance Improvement Method of Convolutional Neural Network Using Combined Parametric Activation Functions)

  • 고영민;이붕항;고선우
    • 정보처리학회논문지:소프트웨어 및 데이터공학
    • /
    • 제11권9호
    • /
    • pp.371-380
    • /
    • 2022
  • 합성곱 신경망은 이미지와 같은 격자 형태로 배열된 데이터를 다루는데 널리 사용되고 있는 신경망이다. 일반적인 합성곱 신경망은 합성곱층과 완전연결층으로 구성되며 각 층은 비선형활성함수를 포함하고 있다. 본 논문은 합성곱 신경망의 성능을 향상시키기 위해 결합된 파라메트릭 활성함수를 제안한다. 결합된 파라메트릭 활성함수는 활성함수의 크기와 위치를 변환시키는 파라미터를 적용한 파라메트릭 활성함수들을 여러 번 더하여 만들어진다. 여러 개의 크기, 위치를 변환하는 파라미터에 따라 다양한 비선형간격을 만들 수 있으며, 파라미터는 주어진 입력데이터에 의해 계산된 손실함수를 최소화하는 방향으로 학습할 수 있다. 결합된 파라메트릭 활성함수를 사용한 합성곱 신경망의 성능을 MNIST, Fashion MNIST, CIFAR10 그리고 CIFAR100 분류문제에 대해 실험한 결과, 다른 활성함수들보다 우수한 성능을 가짐을 확인하였다.

철근 비부착 및 절단 기법을 사용한 PC 접합부의 실험적 연구 (Experimental Study on the Precast Concrete Joint using Re-bar Debonding and Cutting Technique)

  • 이원호;문정호;이용재;이한준
    • 한국방재학회 논문집
    • /
    • 제2권4호
    • /
    • pp.131-141
    • /
    • 2002
  • 지하주차장을 PC화할 때 각각의 부재 접합상태에 따라 설계방법 및 시공방법에 많은 차이가 발생하게 된다. 특히 PC 벽체를 상호 연결하기 위한 접합부에서 수평철근을 겹침이음 방법에 의해 접합하는 경우, 수평철근의 이음길이를 충분히 확보하여야 하기 때문에 시공성이 저하되는 문제점이 있다. 따라서, PC 벽체 수평철근의 돌출길이를 최소화하기 위하여 철근 비부착 기법 및 철근 절단 기법을 적용하는 것을 본 연구의 목적으로 하고 있다. 즉, 철근 콘크리트 부재에서의 위험단면의 위치는 일반적으로 부재의 단부가 되고, 철근의 정착길이는 이 위험단면으로부터의 길이가 된다. 그런데 위험단면의 위치를 비부착 철근 공법 및 철근 절단 공법을 사용하여 부재 안쪽으로 이동함으로서, 수평철근의 정착길이를 최소화시킬 수 있다. 본 연구에서는 철근 비부착 기법 및 철근 절단 기법의 적용 가능성을 파악하기 위해 5개의 실험체를 실물크기로 제작하여 실험적 연구를 실시하고, 절단 철근 및 비부착 철근이 시험체의 거동에 미치는 영향을 비교 및 분석한 결과, 위험단면의 이동을 위하여 철근 비부착 기법 및 철근 절단 기법의 사용이 가능한 것으로 판단된다.

DEVELOPMENT OF PRECISION ATTITUDE DETERMINATION SYSTEM FOR KOMPSAT-2

  • Yoon Jae-Cheol;Shin Dongseok;Lee Hungu;Lee Young-Ran;Lee Hyunjae;Bang Hyo-Choong;Cheon Yee-Jin;Shin Jae-Min;Moon Hong-Youl;Lee Sang-Ryool;Jeun Gab-Ho
    • 한국우주과학회:학술대회논문집(한국우주과학회보)
    • /
    • 한국우주과학회 2004년도 한국우주과학회보 제13권2호
    • /
    • pp.296-299
    • /
    • 2004
  • KARI precision attitude determination system has been developed for high accurate geo-coding of KOMPSAT-2 image. Sensor data from two star trackers and a IRU are used as measurement and dynamic data. Sensor data from star tracker are composed of QUEST and unit vector filter. Filter algorithms consists of extended Kalman filter, unscented Kalman filter, and least square batch filter. The type of sensor data and filter algorithm can be chosen by user options. Estimated parameters are Euler angle from 12000 frame to optical bench frame, gyro drift rate bias, gyro scale factor, misalignment angle of star tracker coordinate frame with respect to optical bench frame, and misalignment angle of gyro coordinate frame with respect to optical bench frame. In particular, ground control point data can be applied for estimating misalignment angle of star tracker coordinate frame. Through the simulation, KPADS is able to satisfy the KOMPSAT-2 mission requirement in which geo-location accuracy of image is 80 m (CE90) without ground control point.

  • PDF

Comparative analysis of commonly used peak calling programs for ChIP-Seq analysis

  • Jeon, Hyeongrin;Lee, Hyunji;Kang, Byunghee;Jang, Insoon;Roh, Tae-Young
    • Genomics & Informatics
    • /
    • 제18권4호
    • /
    • pp.42.1-42.9
    • /
    • 2020
  • Chromatin immunoprecipitation coupled with high-throughput DNA sequencing (ChIP-Seq) is a powerful technology to profile the location of proteins of interest on a whole-genome scale. To identify the enrichment location of proteins, many programs and algorithms have been proposed. However, none of the commonly used peak calling programs could accurately explain the binding features of target proteins detected by ChIP-Seq. Here, publicly available data on 12 histone modifications, including H3K4ac/me1/me2/me3, H3K9ac/me3, H3K27ac/me3, H3K36me3, H3K56ac, and H3K79me1/me2, generated from a human embryonic stem cell line (H1), were profiled with five peak callers (CisGenome, MACS1, MACS2, PeakSeq, and SISSRs). The performance of the peak calling programs was compared in terms of reproducibility between replicates, examination of enriched regions to variable sequencing depths, the specificity-to-noise signal, and sensitivity of peak prediction. There were no major differences among peak callers when analyzing point source histone modifications. The peak calling results from histone modifications with low fidelity, such as H3K4ac, H3K56ac, and H3K79me1/me2, showed low performance in all parameters, which indicates that their peak positions might not be located accurately. Our comparative results could provide a helpful guide to choose a suitable peak calling program for specific histone modifications.

비정상성을 고려한 원평천 유역의 미래 설계홍수량 산정 (Estimation of Future Design Flood Under Non-Stationarity for Wonpyeongcheon Watershed)

  • 류정훈;강문성;박지훈;전상민;송정헌;김계웅;이경도
    • 한국농공학회논문집
    • /
    • 제57권5호
    • /
    • pp.139-152
    • /
    • 2015
  • Along with climate change, it is reported that the scale and frequency of extreme climate events show unstable tendency of increase. Thus, to comprehend the change characteristics of precipitation data, it is needed to consider non-stationary. The main objectives of this study were to estimate future design floods for Wonpyeongcheon watershed based on RCP (Representative Concentration Pathways) scenario. Wonpyeongcheon located in the Keum River watershed was selected as the study area. Historical precipitation data of the past 35 years (1976~2010) were collected from the Jeonju meteorological station. Future precipitation data based on RCP4.5 were also obtained for the period of 2011~2100. Systematic bias between observed and simulated data were corrected using the quantile mapping (QM) method. The parameters for the bias-correction were estimated by non-parametric method. A non-stationary frequency analysis was conducted with moving average method which derives change characteristics of generalized extreme value (GEV) distribution parameters. Design floods for different durations and frequencies were estimated using rational formula. As the result, the GEV parameters (location and scale) showed an upward tendency indicating the increase of quantity and fluctuation of an extreme precipitation in the future. The probable rainfall and design flood based on non-stationarity showed higher values than those of stationarity assumption by 1.2%~54.9% and 3.6%~54.9%, respectively, thus empathizing the necessity of non-stationary frequency analysis. The study findings are expected to be used as a basis to analyze the impacts of climate change and to reconsider the future design criteria of Wonpyeongcheon watershed.

재생냉각 연소기의 냉각기구에 따른 특성 파악 (Investigation of Characteristics for Cooling Parameters of a Combustor in Liquid Rocket Combustors)

  • 김홍집;최환석
    • 한국추진공학회지
    • /
    • 제14권5호
    • /
    • pp.45-50
    • /
    • 2010
  • 연료링의 위치 및 열차폐 코팅의 종류에 따른 연소기의 재생냉각 특성을 검토하였다. 연료링을 노즐의 중간 부분에 위치시키고 냉각채널을 분기시켜서 설계하는 방법이 열적으로 타당함을 확인하였다. 또한 복합재를 이용한 노즐확장부가 적용 가능한 기술적 상황이라면, 팽창비가 높고 열유속이 낮은 노즐 후류 부분은 이를 이용하는 것이 매우 적절하다고 판단된다. 적용 가능한 열차폐 코팅 중에서 30톤급 연소기 및 가스발생기 개발과정에서 사용했던 $Y_2O_3$ stabilized $ZrO_2$과 내산화성이 우수한 Ni/Cr을 고려하였다. 내산화성이 우수한 Ni/Cr에 비해 세라믹 코팅($Y_2O_3$ stabilized $ZrO_2$)이 열차폐 효과가 우수한 것으로 파악되었다.

재생냉각 연소기의 냉각기구에 따른 냉각 특성 파악 (Regenerative Cooling Characteristics for Cooling Parameters of a Combustor in Liquid Rocket Combustors)

  • 김홍집;최환석
    • 한국추진공학회:학술대회논문집
    • /
    • 한국추진공학회 2010년도 제34회 춘계학술대회논문집
    • /
    • pp.145-149
    • /
    • 2010
  • 연료링의 위치 및 열차폐 코팅의 종류에 따른 연소기의 재생냉각 특성을 검토하였다. 연료링을 노즐의 중간 부분에 위치시키고 냉각채널을 분기시켜서 설계하는 방법이 열적으로 타당함을 확인하였다. 또한 복합재를 이용한 노즐확장부가 적용 가능한 기술적 상황이라면, 팽창비가 높고 열유속이 낮은 노즐 후류 부분은 이를 이용하는 것이 매우 적절하다고 판단된다. 적용 가능한 열차폐 코팅 중에서 30톤급 연소기 및 가스발생기 개발과정에서 사용했던 $Y_2O_3$ stabilized $ZrO_2$과 내산화성이 우수한 Ni/Cr을 고려하였다. 내산화성이 우수한 Ni/Cr에 비해 세라믹 코팅($Y_2O_3$ stabilized $ZrO_2$)이 열차폐 효과가 우수한 것으로 파악되었다.

  • PDF

LMI-BASED $H_{\infty}$ LATERAL CONTROL OF AN AUTONOMUS VEHICLE BY LOOK-AHEAD SENSING

  • Kim, C.S.;Kim, S.Y.;Ryu, J.H.;Lee, M.H.
    • International Journal of Automotive Technology
    • /
    • 제7권5호
    • /
    • pp.609-618
    • /
    • 2006
  • This paper presents the lateral control of an autonomous vehicle by using a look-ahead sensing system. In look-ahead sensing by an absolute positioning system, a reference lane, constructed by straight lanes or circular lanes, was switched by a segment switching algorithm. To cope with sensor noise and modeling uncertainty, a robust LMI-based $H_{\infty}$ lateral controller was designed by the feedback of lateral offset and yaw angle error at the vehicle look-ahead. In order to verify the safety and the performance of lateral control, a scaled-down vehicle was developed and the location of the vehicle was detected by using an ultrasonic local positioning system. In the mechatronic scaled-down vehicle, the lateral model and parameters are verified and estimated by a J-turn test. For the lane change and reference lane tracking, the lateral controllers are used experimentally. The experimental results show that the $H_{\infty}$ controller is robust and has better performance compared with look-down sensing.

진동을 이용한 철도차량의 무게중심과 관성모멘트 현장 측정 (Field Measurement of the Center of Gravity and the Moment of Inertia of Railway Vehicles Using Vibration)

  • 송기석;최연선
    • 한국소음진동공학회논문집
    • /
    • 제23권10호
    • /
    • pp.878-884
    • /
    • 2013
  • The center of gravity and the moment of inertia of railway vehicles are important parameters for running safety and stability in railway vehicle design. However, the exact measurement of those is difficult in manufacturing field. The weight measurement of a railway vehicle beneath the wheel using a weight scale is off by a large amount. This paper suggests a measurement method for the center of gravity and the moment of inertia of railway vehicles using vibration. For the measurement a railway vehicle is suspended using four wires. Direct measurement of the tension of the wires and the period of swinging motion of the suspended railway vehicle with calculations give the exact location of the center of gravity and the moment of inertia in x, y, and z directions, respectively. This implementation was demonstrated using an experimental device and verified numerically.

An experimental study on strengthening of vulnerable RC frames with RC wing walls

  • Kaltakci, M. Yasar;Yavuz, Gunnur
    • Structural Engineering and Mechanics
    • /
    • 제41권6호
    • /
    • pp.691-710
    • /
    • 2012
  • One of the most popular and commonly used strengthening techniques to protect against earthquakes is to infill the holes in reinforced concrete (RC) frames with fully reinforced concrete infills. In some cases, windows and door openings are left inside infill walls for architectural or functional reasons during the strengthening of reinforced concrete-framed buildings. However, the seismic performance of multistory, multibay, reinforced concrete frames that are strengthened by reinforced concrete wing walls is not well known. The main purpose of this study is to investigate the experimental behavior of vulnerable multistory, multibay, reinforced concrete frames that were strengthened by introducing wing walls under a lateral load. For this purpose, three 2-story, 2-bay, 1/3-scale test specimens were constructed and tested under reversed cyclic lateral loading. The total shear wall (including the column and wing walls) length and the location of the bent beam bars were the main parameters of the experimental study. According to the test results, the addition of wing walls to reinforced concrete frames provided significantly higher ultimate lateral load strength and higher initial stiffness than the bare frames did. While the total shear wall length was increased, the lateral load carrying capacity and stiffness increased significantly.