• The Sea:JOURNAL OF THE KOREAN SOCIETY OF OCEANOGRAPHY
• /
• v.7 no.3
• /
• pp.100-107
• /
• 2002

The sea-surface winds during the passage of 64 typhoons for 1979-1999 were simulated using two different typhoon wind models, ie, typhoon parametric model(TPM) and primitive vortex model(PVM). The model hindcast winds were compared with the winds observed at JMA ocean buoys(22001 and 21002) and Kyushu ocean observation tower. The analysis of ms and relative errors between hindcast and observed winds was made to find the accuracy and sensitivity of the typhoon wind prediction models. Both hindcast winds of TPM and PVM underestimate the observed typhoon winds, but PVM winds are more closer to the observations with less rms and relative errors. Relative errors of two model winds were small within 200km from typhoon center, but TPM's relative errors increase up to 70% as the radial distance from typhoon center increases beyond > 200km although PVM's relative errors remain in 20% with less sensitive to the distance from typhoon centers.

• International Journal of Aeronautical and Space Sciences
• /
• v.18 no.1
• /
• pp.118-128
• /
• 2017

In this study, we propose a real time inertial navigation system/global positioning system (INS/GPS) integrated smoothing algorithm based on an extended Kalman filter (EKF) and a position damping loop (PDL) for synthetic aperture radar (SAR). Integrated navigation algorithms usually induce discontinuities due to error correction update by the Kalman filter, which are as detrimental to the performance of SAR as the relative position error. The proposed smoothing algorithm suppresses these discontinuities and also reduces the relative position error in real time. An EKF estimates the navigation errors and sensor biases, and all the errors except for the position error are corrected directly and instantly. A PDL activated during SAR operation period imposes damping effects on the position error estimates, where the estimated position error is corrected smoothly and gradually, which contributes to the real time smoothing and small relative position errors. The residual errors are re-estimated by the EKF to maintain the estimation performance and the stability of the overall loop. The performance improvements were confirmed by Monte Carlo simulations. The simulation results showed that the discontinuities were reduced by 99.8% and the relative position error by 48% compared with a conventional EKF without a smoothing loop, thereby satisfying the basic performance requirements for SAR operation. The proposed algorithm may be applicable to low cost SAR systems which use a conventional INS/GPS without changing their hardware configurations.

• Magazine of the Korean Society of Agricultural Engineers
• /
• v.41 no.3
• /
• pp.41-50
• /
• 1999

Derivatio of reasonable design floods was attempted by comparative analysis of design floods derived by Generalized Extreme Value(GEV) distribution using methods of L-moments and LH-moments for the annual maximum series at ten watersheds along Han, Nagdong. Geum, Yeongsan and Seomjin river systems, LH-coefficient of variation, LH-skewness and Lh-kurtosis were calcualted by KH-moment ration respectively. Paramenters were estimated by the Method of LH-Moments, Design floods obtained by Method of LH-Moments using different methods for plotting positionsi n GEV distribution and design floods were compared with those obtained using the Method of L-Moments by the Relative Mean Errors(RME) and Relative Absolute Errors(RAE). The results was found that design floods derived by the method of L-Moments and LH-Moments using Cunnane plotting position formula in the GEV distribution are much closer to those of the observed data in comparison with those obtained by methods of L-moments and LH-moments using the other formula for plotting positions from the viewpoint of Relative Mean Errors and Relative Absolute Errors. In viewpoint of the fact that hydrqulic structures including dams and levees are genrally using design floods with the return period of two hundred years or so, design floods derived by LH-Moments are seemed to be more reasonable than those of L-Moments in the GEV distribution.

• Journal of the Korean Society for Precision Engineering
• /
• v.28 no.8
• /
• pp.904-914
• /
• 2011

The volumetric errors of CNC machining centers are determined by 21 errors, including 3 linear errors, 6 straightness errors, 3 perpendicular errors, 9 angular errors and non-rigid body errors of the machine tool. It is very time consuming and hard to measure all of these errors in which laser interferometer and other parts are used directly. Hence, as many as 21 separate setups and measurements are needed for the linear, straightness, angular and perpendicular errors. In case of the 5-axis machining centers, two more rotary tables are used. It can make 35 error sources of the movement. Therefore, the measured errors of multi movements of the 5-axis tables are very complicated, even if the relative measured errors are measured. This paper describes the methods, those analyze the error sources of the machining centers. Those are based on shifted diagonal measurements method (SDM), R-test and Double ball bar. In case, the angular errors of machine are small enough comparing with others, twelve errors including three linear position errors, six straightness errors and three perpendicular errors can be calculated by using SDM. To confirm the proposed method, SDM was applied to measuring 3 axes of machine tools and compared with directly measurement of each errors. In addition, the methods for measuring relative errors of multi-axis analysis methods using R-test and Double Ball Bar are introduced in this paper.

• Magazine of the Korean Society of Agricultural Engineers
• /
• v.20 no.3
• /
• pp.4739-4749
• /
• 1978

This studies were conducted to derive synthetic unitgraphs and triangular unitgraphs correlated with watershed characteristics which can be used to the estimation and control of flood for the rational development of Agricultural water resources. Derived Synthetic unitgraphs and Triangular unitgraphs can be applied to the ungaged watersheds were compared with average unitgraphs by observed data. Seven small watersheds were selected as studying basins Han, Geum, Nakdong, Yeongsan and Inchon river system. The results summarized for these studies are as follows: 1. Average unitgraphs by observed data and dimensionless unitgraphs for synthesis were derived for all river systems. 2. Peak discharge per unit area of the unitgraph, qp, was derived as qp=10-0.389-0.0424Lg with a high significance. 3. Formulas for the base width of unitgraph of 50 and 75 percent for peak flow for each water systems was adopted as Table 5. 4. The base length of the unitgraph, Tb, in hours in connection with time to peak, Tp, in hours was expressed as Tb =4.3Tp. 5. Peak discharge, Qp, were obtained as Table 6 by the Triangular form to all subwatersheds. 6. Relative errors in the peak discharge of the synthetic unitgraphs showed to be 7.3 percent to the peak of observed average unitgraphs except errors of peak discharge for Yeongsan river system. This indicates that Synthetic unitgraphs for the small watersheds of Han, Geum, Nakdong and Inchon river systems can be applied to the ungaged watersheds. On the other hand, It was confirmed that the accuracy of Instantaneous Unit Hydrograph with only 1.6 percent as relative errors was approaching more closely to the observed average unitgraph than that of synthetic unitgraph with relative errors. 23.9 percent for Yeongsan river system. 7. Errors in the peak discharge of the triangular unitgraph to the observed average unitgraph showed to be 0.6 percent to 7.5 percent which can be regarded as a high precision within the range of 200 to 500$\textrm{km}^2$ in area. On the contrary, application of triangular unitgraph within the range of 200$\textrm{km}^2$ in area has defined as a unsuitable method because of high relative errors, 26.4 percent to 61.6 percent.

• Journal of Astronomy and Space Sciences
• /
• v.26 no.1
• /
• pp.59-74
• /
• 2009

Relative navigation system is presented using GPS measurements from a single-channel global positioning system (GPS) simulator. The objective of this study is to provide the real-time inter-satellite relative positions as well as absolute positions for two formation flying satellites in low earth orbit. To improve the navigation performance, the absolute states are estimated using ion-free GRAPHIC (group and phase ionospheric correction) pseudo-ranges and the relative states are determined using double differential carrier-phase data and singled-differential C/A code data based on the extended Kalman filter and the unscented Kalman filter. Furthermore, pseudo-relative dynamic model and modified relative measurement model are developed. This modified EKF method prevents non-linearity of the measurement model from degrading precision by applying linearization about absolute navigation solutions not about the priori estimates. The LAMBDA method also has been used to improve the relative navigation performance by fixing ambiguities to integers for precise relative navigation. The software-based simulation has been performed and the steady state accuracies of 1 m and 6 mm ($1{\sigma}$ of 3-dimensional difference errors) are achieved for the absolute and relative navigation using EKF for a short baseline leader/follower formation. In addition, the navigation performances are compared for the EKF and the UKF for 10 hours simulation, and relative position errors are mm-level for the two filters showing the similar trends.

• Proceedings of the KSRS Conference
• /
• 2005.10a
• /
• pp.276-279
• /
• 2005

As the Ubiquitous generation approaches, the importance of the sensor data processing is growing. The data approximation scheme, one of the data processing methods, can be the key of sensor data processing, for it is related not only to the lifetime of sensors but also to the size of the storage. In this paper, we propose the Harmonic Wavelet transform which can minimize the relative error for given sensor data. Harmonic Wavelets use the harmonic mean as a representative which is the minimum point of the maximum relative error between two data values. In addition, Harmonic Wavelets retain the relative errors as wavelet coefficients so we can select proper wavelet coefficients that reduce the relative error more easily. We also adapt the greedy algorithm for local optimization to reduce the time complexity. Experimental results show the performance and the scalability of Harmonic Wavelets for sensor data.

• Atmosphere
• /
• v.29 no.5
• /
• pp.551-566
• /
• 2019

Quantitative understanding of a random error that is associated with Lagrangian particle dispersion modeling is a prerequisite for backward-in-time mode simulations. This study aims to quantify the random error of the WRF-FLEXPART model and suggest an optimum number of the Lagrangian particles for backward-in-time simulations over the Seoul metropolitan area. A series of backward-in-time simulations of the WRF-FLEXPART model has conducted at two receptor points by changing the number of Lagrangian particles and the relative error, as a quantitative indicator of random error, is analyzed to determine the optimum number of the release particles. The results show that in the Seoul metropolitan area a 1-day Lagrangian transport contributes 80~90% in residence time and ~100% in atmospheric enhancement of carbon monoxide. The relative errors in both the residence time and the atmospheric concentration enhancement are larger when the particles release in the daytime than in the nighttime, and in the inland area than in the coastal area. The sensitivity simulations reveal that the relative errors decrease with increasing the number of Lagrangian particles. The use of small number of Lagrangian particles caused significant random errors, which is attributed to the random number sampling process. For the particle number of 6000, the relative error in the atmospheric concentration enhancement is estimated as -6% ± 10% with reduction of computational time to 21% ± 7% on average. This study emphasizes the importance of quantitative analyses of the random errors in interpreting backward-in-time simulations of the WRF-FLEXPART model and in determining the number of Lagrangian particles as well.

• Cross-Cultural Studies
• /
• v.19
• /
• pp.359-388
• /
• 2010

This study is aimed to investigate the aspect of using relative clauses in Korean. The data used for this study were extracted from the Korean text books for the foreign students and from the Chinese students' Korean compositions. They are the learners of Korean language at the early intermediated stage. As the result of analyzing them, the followings were found I)the majority of relative clauses in Korean consists of left-branching sentences. ii)The number of the subject relative clauses was higher than object ones in both of them. Especially in the aspect of using relative clauses, subject ones were used even more frequently than objective ones. This result is corresponded to the previous thesis, "the subject relative clauses was acquired earlier than object ones". iii)The relative clauses that those with a head noun function as subject in the main sentence showed in higher proportion in comparison of those as object. That is, this study showed that subjects were used more frequently than objects in the relative clauses used in their compositions. Finally, this study analyzed the errors of adnominal ending usage occurring in their compositions. More errors occurred when adjective form ended with '-hada' are changed into adnominal ending one.

• Proceedings of the Korean Society of Agricultural Engineers Conference
• /
• 1999.10c
• /
• pp.479-485
• /
• 1999

This study was conducted to derived design floods by Generalized Extreme Value(GEV) distributiion for the annual maximum series at ten watersheds along Han, Nagdong, Geum , Yeongsan and Seomjin river systems. Adequency for the analysis of flood data used in this study was established by the test of Independence, Homogeneity , detection of Outliers. Coefficient of variation , skewness and kurtosis were calculated by the L-Moment, and LH-Moment ratio respectively. Parameters were estimated by the Method of L-Method of LH-Moment. Design floods obtained by Method of L-Moments and LH-Moments using different methods for plotting positions in GEV distributions and were compared with those obatined using the Method of L-Moments and LH-Moments by the Relative Mean Errors and Realtive Absoulte Errors. It was found that desgin floods derived by the method of L-Moments and LH-Moments using Cunnane plotting position foumula in the GEV distribution are much closer to those of the observed data in comparison with those obtained by methods of L-moments and LH-moments using the other formula for poltting postions from the viewpoint of Relative Mean Errors and Relative Absoulte Errors. In view of the fact that hydraulic structures indcluding dams and levees are generally usiong design floods with the return period of two hundred years or so, design floods derived by LH-Moments are seemed to be more reasonable than those of L-Moments in the GEV distribution.