• Title/Summary/Keyword: determination of model parameters

검색결과 510건 처리시간 0.046초

저류함수모형의 매개변수 보정 및 추정 (Calibration and Estimation of Parameter for Storage Function Model)

  • 김범준;곽재원;이진희;김형수
    • 대한토목학회논문집
    • /
    • 제28권1B호
    • /
    • pp.21-32
    • /
    • 2008
  • 홍수예보는 홍수로 인한 인명과 재산피해를 저감하기 위한 비구조적 대책으로 그 정확성은 매우 중요한 요소이다. 그러나 우리나라 홍수예보에서 주로 쓰이는 저류함수모형은 그 적용에 있어서 매개변수의 결정이 매우 중요하지만 이를 결정하는 것은 매우 어렵다. 그래서 현재에는 주로 경험식을 이용하거나 수문 기술자의 판단에 의해 보정이 이루어지고 있는 실정이다. 이에 본 연구에서는 저류함수모형 매개변수의 특성을 알고 범위를 설정하기 위해서 민감도 분석을 수행하였으며, 보정을 위해 다양한 보정방법과 목적함수를 적용하여 그 성능을 평가하였다. 또한 유역에 맞는 매개변수를 결정하기 위한 방법을 제시하고, 결정된 매개변수를 이용하여 검증홍수사상에 대해 검증을 수행하였다. 본 연구의 결과와 기존에 제시된 매개변수를 사용한 결과를 비교하였다. 기존에 제시된 대상유역의 저류함수모형 매개변수도 좋은 결과를 나타내었으나, 새롭게 추정된 매개변수는 보다 더 좋은 결과를 도출하였다.

열화상 이미지와 환경변수를 이용한 콘크리트 균열 깊이 예측 머신 러닝 분석 (Comparison Analysis of Machine Learning for Concrete Crack Depths Prediction Using Thermal Image and Environmental Parameters)

  • 김지형;장아름;박민재;주영규
    • 한국공간구조학회논문집
    • /
    • 제21권2호
    • /
    • pp.99-110
    • /
    • 2021
  • This study presents the estimation of crack depth by analyzing temperatures extracted from thermal images and environmental parameters such as air temperature, air humidity, illumination. The statistics of all acquired features and the correlation coefficient among thermal images and environmental parameters are presented. The concrete crack depths were predicted by four different machine learning models: Multi-Layer Perceptron (MLP), Random Forest (RF), Gradient Boosting (GB), and AdaBoost (AB). The machine learning algorithms are validated by the coefficient of determination, accuracy, and Mean Absolute Percentage Error (MAPE). The AB model had a great performance among the four models due to the non-linearity of features and weak learner aggregation with weights on misclassified data. The maximum depth 11 of the base estimator in the AB model is efficient with high performance with 97.6% of accuracy and 0.07% of MAPE. Feature importances, permutation importance, and partial dependence are analyzed in the AB model. The results show that the marginal effect of air humidity, crack depth, and crack temperature in order is higher than that of the others.

Applying methane and carbon flow balances for determination of first-order landfill gas model parameters

  • Park, Jin-Kyu;Chong, Yong-Gil;Tameda, Kazuo;Lee, Nam-Hoon
    • Environmental Engineering Research
    • /
    • 제25권3호
    • /
    • pp.374-383
    • /
    • 2020
  • Landfill gas (LFG) emissions from a given amount of landfill waste depend on the carbon flows in the waste. The objective of this study was to more accurately estimate the first-order decay parameters through methane (CH4) and carbon flow balances based on the analysis of a full-scale landfill with long-term data and detailed field records on LFG and leachate. The carbon storage factor for the case-study landfill was 0.055 g-degradable organic carbon (DOC) stored per g-wet waste and the amounts of DOC lost with the leachate were less than 1.3%. The appropriate CH4 generation rate constant (k) for bulk waste was 0.24 y-1. The the CH4 generation potential (L0) values ranged 33.7-46.7 m3-CH4 Mg-1, based on the fraction of DOC that can decompose (DOCf) value of 0.40. Results show that CH4 and carbon flow balance methods can be used to estimate model parameters appropriately and to predict long-term carbon emissions from landfills.

지진파(地震波)를 이용(利用)한 남한(南韓)의 지각구조(地殼構造) 연구(硏究) (A Study on the Crustal Structure of South Korea by using Seismic Waves)

  • 김상조;김소구
    • 자원환경지질
    • /
    • 제16권1호
    • /
    • pp.51-61
    • /
    • 1983
  • By using local earthquake data, the Korean crust model and travel-time tables were determined. The upper crustal earthquakes (Hongsung event and Ssanggyesa event) were considered as auxiliary information, and the lower crustal earthquakes (Uljin event and Pohang event) played an important role in determining model parameters. The possible existence of Low Velocity Layer (LVL) in the upper mantle was suggested by discrepancy in the arrival times of Sariwon earthquake which occurred below Moho discontinuity. Computer program for the determination of the model parameters was developed in order to screened out the optimum parameters by comparing the travel times of observed data with theoretical ones. We found that the discontinuities of Conrad, Moho, and upper and lower boundaries of LVL have their depth of 15, 32, 55 and 75 Km, respectively. The velocities of P-and S-wave in the layers between those discontinities were found to be (1) 5.98, 3.40 Km/sec (2) 6.38, 3.79 Km/sec (3) 7.95, 4.58 Km/sec (4) unknown (5) 8.73, 5.05 Km/sec, respectively from the top layer. Travel-time tables were also computed for the inter-local earthquakes which have their direct wave paths above the LVL.

  • PDF

강우-유출모형을 위한 매개변수 순차 보정기법 연구 (A Study of Progressive Parameter Calibrations for Rainfall-Runoff Models)

  • 곽재원;김덕길;홍일표;김형수
    • 한국습지학회지
    • /
    • 제11권2호
    • /
    • pp.107-121
    • /
    • 2009
  • 현재 홍수예보를 위하여 많은 강우-유출 모형이 사용되고 있으나, 이러한 모형의 매개변수를 결정하는 것은 매우 난해하다. 본 연구에서는 저류함수모형과 Tank 모형, SSARR 모형을 이용하여 미호천 유역에 대하여 홍수모의 예측을 수행하고 그 효율성을 분석하였다. 연구에 적용된 강우-유출 모형에 최적화 방법을 적용하여 매개 변수 산정을 수행하였으며, 패턴탐색과 유전자 알고리즘의 최적화 방법을 적용 시, 보정과정 내에서 매개변수 간 민감도를 분석하고 이를 바탕으로 매개변수를 소군집으로 분류하여 민감도에 따른 순차 보정 방법을 적용하고 이 결과를 비교 분석하였다. 매개변수 소군집을 이용한 보정 방법과 기존에 사용되는 매개변수 군집을 이용한 보정 방법을 비교한 결과, SSR에 소군집을 이용한 순차보정 방법을 적용하였을 때 첨두 유량과 보정 시간 면에서 유리한 것으로 나타났다.

  • PDF

Serviceability assessment of subway induced vibration of a frame structure using FEM

  • Ling, Yuhong;Gu, Jingxin;Yang, T.Y.;Liu, Rui;Huang, Yeming
    • Structural Engineering and Mechanics
    • /
    • 제71권2호
    • /
    • pp.131-138
    • /
    • 2019
  • It is necessary to predict subway induced vibration if a new subway is to be built. To obtain the vibration response reliably, a three-dimensional (3D) FEM model, consisting of the tunnel, the soil, the subway load and the building above, is established in MIDAS GTS NX. For this study, it is a six-story frame structure built above line 3 of Guangzhou metro. The entire modeling process is described in detail, including the simplification of the carriage load and the determination of model parameters. Vibration measurements have been performed on the site of the building and the model is verified with the collected data. The predicted and measured vibration response are used together to assess vibration level due to the subway traffic in the building. The No.1 building can meet work and residence comfort requirement. This study demonstrates the applicability of the numerical train-tunnel-soil-structure model for the serviceability assessment of subway induced vibration and aims to provide practical references for engineering applications.

수용액중의 진주층에 대한 염기성 염료의 흡착매개변수 및 흡착모델 선정 (Selection of Adsorption Model and Parameters for Basic Dyes from Aqueous Solution onto Pearl Layer)

  • 신춘환;송동익
    • 한국환경과학회지
    • /
    • 제14권12호
    • /
    • pp.1203-1209
    • /
    • 2005
  • Basic dyes, Rhodamine 6G(R6G), Rhodamine B(RB), and Methylene Blue(MB), dissolved in water were used to investigate single-component adsorption affinity to the pearl layer fractionated according to the size. Unfractionated pearl layers were also used as adsorbents for the R6G and RB. The Langmuir and the Redlich-Peterson(RP) models were used to fit the adsorption data, and the goodness of fit was examined by using determination coefficient($R^2$) and standard deviation(SSE). The 3-parameter RP model was found to be better in describing the dye adsorption data than the 2 parameter Langmuir model, as can be expected from the number of parameters involved in the model. The adsorption affinity to the fractionated pearl layer was higher than that to the unfractionated layer The affinity order to the fractionated Conchiolin layer was found to be R6G > MB > RB. Furthermore, the dye adsorption capacity of the various types of pearl layer was found to be in the order, the fractionated pearl > powdered pearl > unfractionated pearl, exhibiting different adsorption isotherms according to the types of layer used in the study.

계통보호릴레이와의 협조를 고려한 SFCL의 동작파라메타 설정방법에 대한고찰 (On Setting Method of the operating Parameters of SFCL in Transmission Systems Considering Power Protection Relay)

  • 홍원표
    • 대한전기학회:학술대회논문집
    • /
    • 대한전기학회 1998년도 하계학술대회 논문집 C
    • /
    • pp.1231-1234
    • /
    • 1998
  • Design & Operation of power system for meeting increase of electric power demand is becoming more difficult and complex. One of reasons is increase of fault current. As one of the most effective methods for suppressing the fault current, installation of SFCL is expected. This paper describes a method of fault analyses of power system with SFCLs, and also discusses determination of specification of SFCLs, effects of limiting the fault current due to SFCLs by use of the model system of two - bus electric power system with parallel circuit model transmission line. Also, describes the definition of six specific parameters of SFCL for power system application & a proposal of design method of specific parameter of a resistance type SFCL in overhead transmission lines considering operation of protective relays.

  • PDF

Photogrammetric Modeling of KOMPSAT Stereo Strips Using Minimum Control

  • Yoo, Hwan-Hee;Sohn, Hong-Gyoo;Kim, Seong-Sam;Jueng, Joo-Kweon
    • Korean Journal of Geomatics
    • /
    • 제2권1호
    • /
    • pp.31-35
    • /
    • 2002
  • This paper describes an experiment for three-dimensional positioning for a pair of KOMPSAT stereostrips using the ancillary data and a single ground control point. The photogrammetric model for three-dimensional positioning was performed as follows: first, initialization of orbital and attitude parameters derived from ancillary data; second, adjustment of orbital and attitude parameters for the satellite to minimize the ground position error with respect to a GCP using the collinearity condition; third, determination of actual satellite position; and lastly, space intersection. This model was tested for a pair of stereo strips with 0.6 base-to-height ratio and GCPs identified from a 1:5,000 scale digital map. As the result, the satellite position of offset was corrected by only one GCP and the accuracy for the geometric modeling showed 38.89m RMSE.

  • PDF

Coronary Artery Numerical Flow Analysis for Determination of Bypass Graft Geometric Parameters

  • Kim, Hyung-Min;Kim, Woong
    • Journal of Mechanical Science and Technology
    • /
    • 제19권3호
    • /
    • pp.905-912
    • /
    • 2005
  • A computational investigation of blood flow in a coronary artery grafted by artificial bypass was performed to determine such geometric parameters as the curvature of radius, approach length, and angle of end-to-side anastomosis. Transient flow features in the host artery were computed using FVM and SIMPLE algorithms. We compared flow distributions and wall shear stresses in two simple models, planar and non-planar, and confirmed that the non-planar bypass model was more conducive to suppressing intimal hyperplasia. Our non-planar model with $60^{\circ}$ of anastomosis and a 1.0 diameter approach length and radius of curvature predicts a relatively small, spatially-extended high-OSI (>0.01) zone, as well as an increased average wall shear stress on this zone.