• 한국수자원학회논문집
• /
• 제35권5호
• /
• pp.575-581
• /
• 2002

본 연구에서는 강수의 계절성에 따라 면적평균강수의 추정오차가 어떻게 달라지는지를 평가하였다. 공간상관을 고려하는 경우와 고려하지 않는 경우 모두를 다루었으며, 각각의 경우에 대해 추정오차의 변화를 살펴보았다. 유사한 경우로서 계절성을 무시하고 누가시간을 증가시켜 추정오차가 어떻게 변하는지도 살펴보았다. 본 연구는 금강유역에 적용하였으며 30년 이상의 일 강수 기록을 가진 28개 지점의 자료를 이용하였다. 본 연구의 결과를 정리하면 다음과 같다: (1) 월 단위의 면적평균강수량에 대한 추정오차는 대체로 강수량에 비례하여 나타나며, 따라서 강한 계절성을 나타낸다. 그러나 이를 평균 강수량으로 나눈 상대오차는 1월과 12월을 제외하면 대략 5 - 8% 정도로 유사한 값을 보인다. (2) 연 강수량에 대한 추정오차는 연강수량의 3% 정도인 것으로 나타났다. (3) 그러나, 강수량이 아닌 강수량의 표준편차를 기준으로 삼는 경우 면적평균강우의 추정오차는 월 단위 및 년 단위에서 동일하게 표준편차의 11% 정도로 계산된다. (4) 마지막으로, 공간상관을 고려하지 않는 경우의 추정오차는 고려하는 경우의 2배 정도까지 커짐을 확인할 수 있었다.

• 대한토목학회논문집
• /
• 제31권1B호
• /
• pp.21-27
• /
• 2011

유출예측량을 모의할 때 과거와 현재의 수문자료를 이용한다는 측면에서 미래 예측결과의 불확실성을 완전히 제거할 수는 없겠지만, 다양한 기법별 분석에 의하여 불확실성을 감소시킬 수 있다. 본 연구에서는 유출예측의 정확성 향상을 위해 다양한 유출예측 기법을 적용 및 평가하였으며 확률론적 예측을 가능하게 하는 예측기법인 ESP와 관측 시계열 자료를 이용한 통계기법으로 공주지점의 유출예측을 수행하였다. 각 기법에 따른 유출예측 결과의 신뢰성 평가는 MAE(Mean Absolute Error), RMSE(Root Mean Squared Error), RRMSE(Relative Root Mean Squared Error), Mean Absolute Percentage Error (MAPE), TIC(Theil Inequality Coefficient)를 이용하였다. ESP 확률을 이용하여 예측한 유출결과와 통계적 시계열 분석에 의해 예측된 유출결과를 MAE, RMSE, RRMSE, MAPE, TIC를 이용하여 비교 분석하였으며 유출예측의 개선효과를 확인해본 결과, ESP 확률을 이용한 예측이 MAE(10.6), RMSE(15.14), RRMSE(0.244), MAPE(22.74%), TIC(0.13)으로 평가되었으며 MAE(23.2), RMSE(37.13), RRMSE(0.596), MAPE(26.69%), TIC(0.30)으로 평가된 ARMA와 MAE(26.4), RMSE(34.44), RRMSE(0.563), MAPE(47.38%), TIC(0.25)으로 평가된 Winters 에 비해 신뢰성이 높게 나타났다.

• 한국방재학회:학술대회논문집
• /
• 한국방재학회 2008년도 정기총회 및 학술발표대회
• /
• pp.97-100
• /
• 2008

This study estimated the error involved in the areal average rainfall derived incomplete radar information due to radar partial coverage of a basin or sub-basin. This study considers the Han River Basin as an application example for the rainfall observation using the Ganghwa rain radar. Among the total of 24 mid-sized sub-basins of the Han River Basin evaluated in this study, only five sub-basins are fully covered by the radar and three are totally uncovered. Remaining 16 sub-basins are partially covered by the radar leading incomplete radar information available. When only partial radar information is available, the sampling error decreases proportional to the size of the radar coverage, which also varies depending on the number of clusters. It is general that smaller sampling error can be expected when the number of clusters increases if the total area coverage remains the same. This study estimated the sampling error of the areal average rainfall of partially-covered mid-sized sub-basins of the Han River Basin, and the results show that the sampling error could be at least several % to maximum tens % depending on the relative coverage area.

• 수산해양기술연구
• /
• 제35권3호
• /
• pp.215-226
• /
• 1999

The experimented rangefinder consist of sets of V/A-Code GPS and sets of L1 C/A-code & carrier phase receivers connected by two spread spectrum radio modems in order to measure relative range and bearing between two ship antennas by real time, comparing and analyzing accuracy of both GPS receivers at the fix point on the land by means of executing zero baseline test by C/A code and by carrier phase as well as measuring distance range 5m, 10m, 15m between each other receivers. The results from the measurement of relative range and bearing are as follows as ;1. According to the results from zero baseline test, the average error by C/A-code receiver is less than 0.1m, which proves theories from published books but when each GPS receivers track different satellites, the range accuracy error becomes up to 100m by means of S/A. Because of this sudden wide range error, rangefinder is not appropriate at relative range measurement without additional modification of the algorism of the GPS receiver itself.2. According to relative range measurement by Carrier Phase and zero baseline test at static condition, the range error is less than 3.5cm in case that it passes more than 5 minutes after GPS sets can track simultaneously more than 6 satellites. Its main reason is understood that the phase center of antenna is bigger than geodetic antenna.3. When range measurement of two receivers from 5m, to 10m to 15m, the each range error is 0.340m, 0.190m, 0.011m and each standard variation is 0.0973m, 0.0884m, 0.0790m. The range error and standard variation are in inverse proportion to distance between two receivers. 4. L1 Carrier Phase GPS generally needs 5 minutes to fix and during this ambiguity search, the relative range and bearing angle is shown to be various.

• ETRI Journal
• /
• 제34권5호
• /
• pp.696-705
• /
• 2012

Accurate synthesis models based on artificial neural networks (ANNs) are proposed to directly obtain the physical dimensions of an asymmetric coplanar waveguide with conductor backing and substrate overlaying (ACPWCBSO). First, the ACPWCBSO is analyzed with the conformal mapping technique (CMT) to obtain the training data. Then, a modified genetic-algorithm-Levenberg-Marquardt (GA-LM) algorithm is adopted to train ANNs. In the algorithm, the maximal relative error (MRE) is used as the fitness function of the chromosomes to guarantee that the MRE is small, while the mean square error is used as the error function in LM training to ensure that the average relative error is small. The MRE of ANNs trained with the modified GA-LM algorithm is less than 8.1%, which is smaller than those trained with the existing GA-LM algorithm and the LM algorithm (greater than 15%). Lastly, the ANN synthesis models are validated by the CMT analysis, electromagnetic simulation, and measurements.

• Journal of applied mathematics & informatics
• /
• 제15권1_2호
• /
• pp.363-375
• /
• 2004

In this paper, We show that the bandpass random signals of the form ∑$_{\alpha}$$\alpha$$_{\alpha}$ a Sin(2$\pi$f$_{\alpha}$t + b$_{\alpha}$) where a$_{\alpha}$ being a random number in [0,1], f$_{\alpha}$ a random integer in a given frequency band, and b$_{\alpha}$ a random number in [0, 2$\pi$], generate Gaussian white noise signals and hence they are adequate for simulating Continuous Markov processes. We apply the result to the fluctuation-dissipation formula for the Johnson noise and show that the probability distribution for the long term average of the power of the Johnson noise is a X$^2$ distribution and that the relative error of the long term average is (equation omitted) where N is the number of blocks used in the average.error of the long term average is (equation omitted) where N is the number of blocks used in the average.

• Journal of Advanced Marine Engineering and Technology
• /
• 제38권2호
• /
• pp.188-194
• /
• 2014

Recently, indoor localization systems based on wireless sensor networks have received a great deal of attention because they help achieve high accuracy in position determination by using various algorithms. In order to minimize the error in the estimated azimuth that can occur owing to sensor drift and recursive calculation in these algorithms, we propose a novel relative azimuth estimation algorithm. The advantages of the proposed technique in an indoor environment are that an improved weight average filter is used to effectively reduce impulse noise from the raw data acquired from nodes with inherent errors and a rotational displacement algorithm is applied to obtain a precise relative azimuth without using additional sensors, which can be affected by electromagnetic noise. Results from simulations show that the proposed filter reduces the impulse noise, and the acquired estimation error does not accumulate with time by using proposed algorithm.

• Current Optics and Photonics
• /
• 제2권4호
• /
• pp.348-355
• /
• 2018

How to measure the creepage distance of an insulator quickly and accurately is a problem for the power industry at present, and the noticeable concern is that the high voltage insulation equipment cannot be measured online in the charged state. In view of this situation, we develop an on-line measurement system of creepage distance for high voltage insulators based on binocular stereo vision. We have proposed a method of generating linear structured light using a conical off-axis mirror. The feasibility and effect of two ways to solve the interference problem of strong sunlight have been discussed, one way is to use bandpass filters to enhance the contrast ratio of linear structured light in the images, and the other way is to process the images with adaptive threshold segmentation and feature point extraction. After the system is calibrated, we tested the measurement error of the on-line measurement system with a composite insulator sample. Experimental results show that the maximum relative error is 1.45% and the average relative error is 0.69%, which satisfies the task requirement of not more than 5% of the maximum relative error.

• KSII Transactions on Internet and Information Systems (TIIS)
• /
• 제9권10호
• /
• pp.3983-4001
• /
• 2015

Exact average bit error probability (BEP) expressions for mobile-relay-based mobile-to-mobile (M2M) cooperative networks with incremental decode-and-forward (IDF) relaying over N-Nakagami fading channels are derived in this paper. The average BEP performance under different conditions is evaluated numerically to confirm the accuracy of the analysis. Results are presented which show that the fading coefficient, the number of cascaded components, the relative geometrical gain, and the power allocation parameter have a significant influence on the average BEP performance.

• 대한임베디드공학회논문지
• /
• 제15권5호
• /
• pp.235-242
• /
• 2020

This paper analyzes the effect of reduced output width of the truncated logarithmic multiplication and application to inferences using convolutional neural networks (CNNs). For small hardware overhead, output width is reduced in the truncated Mitchell multiplier, so that fractional bits in multiplication output are minimized in error-resilient applications. This analysis shows that when reducing output width in the truncated Mitchell multiplier, even though worst-case relative error increases, average relative error can be kept small. When adopting 8 fractional bits in multiplication output in the evaluations, there is no significant performance degradation in target CNNs compared to existing exact and original Mitchell multipliers.