• Journal of Broadcast Engineering
• /
• v.27 no.2
• /
• pp.174-184
• /
• 2022

3D human pose estimation models can be classified into a multi-view model and a single-view model. In general, the multi-view model shows superior pose estimation performance compared to the single-view model. In the case of the single-view model, the improvement of the 3D pose estimation performance requires a large amount of training data. However, it is not easy to obtain annotations for training 3D pose estimation models. To address this problem, we propose a method to generate pseudo ground-truths of multi-view human pose data from a multi-view model and exploit the resultant pseudo ground-truths to train a single-view model. In addition, we propose a multi-view consistency loss function that considers the consistency of poses estimated from multi-view images, showing that the proposed loss helps the effective training of single-view models. Experiments using Human3.6M and MPI-INF-3DHP datasets show that the proposed method is effective for training single-view 3D human pose estimation models.

• Journal of the Earthquake Engineering Society of Korea
• /
• v.10 no.3 s.49
• /
• pp.101-111
• /
• 2006

This paper proposes a method for modeling new fault ground motion due to moderate size earthquakes in Stable Continental Regions (SCRs) for the first time. The near fault ground motion is characterized by a single long period velocity pulse of large amplitude. In order to model the velocity pulse, its period and peak amplitude need be determined in terms of earthquake magnitude and distance from the causative fault. Because there have been observed very few new fault ground motions, it is difficult to derive the model directly from the recorded data in SCRs. Instead an indirect approach is adopted in this work. The two parameters, the period and peak amplitude of the velocity pulse, are known to be functions of the rise time and the slip velocity. For Western United States (WUS) that belongs active tectonic regions, there art empirical formulas for these functions. The relations of rise time and slip velocity on the magnitude in SCRs are derived by comparing related data between Western United States and Central-Eastern United States that belongs to SCRs. From these relations, the functions of these pulse parameters for NFGM in SCRs can be expressed in terms of earthquake magnitude and distance. A time history of near fault ground motion of moderate magnitude earthquake in stable continental regions is synthesized by superposing the velocity pulse on the for field ground motion that is generated by stochastic method. As an demonstrative application, the response of a single degree of freedom elasto-plastic system is studied.

• Earthquakes and Structures
• /
• v.10 no.6
• /
• pp.1315-1330
• /
• 2016

This study investigates ground motion parameters and their damage potential for building type structures. It focuses on low and mid-rise reinforced concrete buildings that are important portion of the existing building stock under seismic risk in many countries. Correlations of 19 parameters of 466 earthquake records with nonlinear displacement demands of 1056 Single Degree of Freedom (SDOF) systems are investigated. Properties of SDOF systems are established to represent RC building construction practice. The correlation of damage and ground motion characteristics is examined with respect to number of story and site classes. Equations for average nonlinear displacement demands of considered RC buildings are given for some of the ground motion parameters. Velocity related parameters are generally found to have better results than the acceleration, displacement and frequency related ones. Correlation of the parameters may be expected to decrease with increasing intensity of seismic event. Velocity Spectrum Intensity and Peak Ground Velocity have been found to have the highest correlation values for almost all site classes and number of story groups. Common parameter of Peak Ground Acceleration has lower correlation with damage when compared to them and some other parameters like Effective Design Acceleration and Characteristic Intensity.

• Journal of Electrical Engineering and Technology
• /
• v.6 no.1
• /
• pp.1-7
• /
• 2011

This paper proposes an enhanced noise robust algorithm for fault location on double-circuit transmission line for the case of single line-to-ground (SLG) fault, which uses distributed parameter line model that also considers the mutual coupling effect. The proposed algorithm requires the voltages and currents from single-terminal data only and does not require adjacent circuit current data. The fault distance can be simply determined by solving a second-order polynomial equation, which is achieved directly through the analysis of the circuit. The algorithm, which employs the faulted phase network and zero-sequence network with source impedance involved, effectively eliminates the effect of load flow and fault resistance on the accuracy of fault location. The proposed algorithm is tested using MATLAB/Simulink under different fault locations and shows high accuracy. The uncertainty of source impedance and the measurement errors are also included in the simulation and shows that the algorithm has high robustness.

• Earthquakes and Structures
• /
• v.11 no.5
• /
• pp.905-924
• /
• 2016

Several two-dimensional analytical beam column joint models with varying complexities have been proposed in quantifying joint flexibility during seismic vulnerability assessment of non-ductile reinforced concrete (RC) frames. Notable models are the single component rotational spring element and the super element joint model that can effectively capture the governing inelastic mechanisms under severe ground motions. Even though both models have been extensively calibrated and verified using quasi-static test of joint sub-assemblages, a comparative study of the inelastic seismic responses under nonlinear time history analysis (NTHA) of RC frames has not been thoroughly evaluated. This study employs three hypothetical case study RC frames subjected to increasing ground motion intensities to study their inherent variations. Results indicate that the super element joint model overestimates the transient drift ratio at the first story and becomes highly un-conservative by under-predicting the drift ratios at the roof level when compared to the single-component model and the conventional rigid joint assumption. In addition, between these story levels, a decline in the drift ratios is observed as the story level increased. However, from this limited study, there is no consistent evidence to suggest that care should be taken in selecting either a single or multi component joint model for seismic risk assessment of buildings when a global demand measure such as maximum inter-storey drift is employed in the seismic assessment framework.

• Proceedings of the Korean Institute of Electrical and Electronic Material Engineers Conference
• /
• 2005.11a
• /
• pp.157-158
• /
• 2005

Superconducting fault currents(SFCLs) are expected to improve not only reliability but also stability of real power systems. The analysis on the single line-to-ground fault of the integrated three phase flux-lock type SFCL, which consists of three flux-lock reactor wound on an iron core in each single phase and three YBCO thin films, was investigated in current limiting operation characteristics. We compared 21turn numbers with 42turn numbers according to wound turn numbers each the coil 2 under the additive polarity winding operation between coil 1 and coil 2. The three phase flux-lock type SFCL using an iron core differently operates general three phase resistive SFCLs. When a single line-to-ground fault occurred, the SFCL's three units were quenched after fault onset. We confirmed effective current limiting operation characteristics with adjustable inductance level.

• Geotechnical Engineering
• /
• v.3 no.3
• /
• pp.21-30
• /
• 1987

Some methods are presented to estimate the ultimate lateral resistance of single active piles subjected to lateral loads above the ground surface, considering the lateral soil reaction, the pile length and the fixity condition of a pile head. The lateral soil reaction acting on a single active pile embedded in soil due to pile movement can be estimated by use of a theoretical equation which is derived by considering especially the plastic state of ground surrounding the pile and the pile- section shape. The piles are named short or long depending upon the relative magnitude of the induced bending moment to the yielding moment. As for the fixity condition of a pile head, the free head and the unrotated head are considered. Comparison with other experimental results gives that the calculated ultimate lateral resistance obtained by the author's theory is closer to experimental results than the one obtained by Brom's theory.

• The Transactions of the Korean Institute of Power Electronics
• /
• v.23 no.2
• /
• pp.129-138
• /
• 2018

Previous single-phase transformerless inverters used active bypass switching circuits that need auxiliary power switches to minimize ground leakage current. However, switching and gate driving losses are increased due to the use of additional power switches. To cope with this drawback, this work proposes a transformerless inverter using a passive bypass filter without any auxiliary power switch. The operation and control of the proposed inverter are described. The ground leakage current characteristics are analyzed for the proposed inverter with the passive bypass filter. The experimental results of the proposed inverter for a 1.0kW prototype system are presented.

• Journal of the Korean Institute of Illuminating and Electrical Installation Engineers
• /
• v.20 no.9
• /
• pp.70-77
• /
• 2006

We have fabricated an integrated three-phase flux-lock type SFCL, which consists of an YBCO($YB_a2Cu_3O_7$) thin film and a flux-lock reactor wound around an iron core of each phase. In order to apply the SFCL in a real power system, fault analyses for the three-phase system are essential. The short-circuit currents were effectively limited by adjusting the numbers of winding of each secondary coil and their winding directions. The flux flow generated in the iron core cancelled out under the normal operation due to the parallel connection between primary and secondary windings. However, the flux-lock type SFCL with same iron core was operated just after the fault due to the flux generating in the iron core. To analyze the current limiting characteristics, the additive polarity winding was compared with the subtractive one in the flux lock reactor. Whenever a single line-to-ground fault occurred in any phase, the peak value of the line current of the fault phase in the additive polarity winding increased up to about 12.87 times during the first-half cycle. On the other hand, the peak value in the subtractive polarity winding increased up to about 34.07 times under the same conditions. This is because the current flow between the primary and the secondary windings changed to additive or subtractive status according to the winding direction. We confirmed that the current limiting behavior in the additive polarity winding was more effective for a single-line-to-ground fault

• Journal of the Korean Geotechnical Society
• /
• v.36 no.1
• /
• pp.5-15
• /
• 2020

Dynamic behavior of pile foundation is significantly influenced by the dynamic interaction between soil and pile. Especially, in the sloping ground, the soil-pile interaction becomes very complex due to different resistance according to loading direction, soil residual displacement and so on. In this study, dynamic centrifuge tests were performed on the piles in the sloping ground. The model structures consisted of a single pile and 2×2 group pile. The soil-pile interaction has been investigated considering various conditions such as slope, single and group piles, and amplitude of input motions. The phase differences between soil and pile displacement and dynamic p-y curves were evaluated. The analysis results showed that the pile behavior was largely influenced by the kinematic forces between soil and pile. In addition, the dynamic p-y curve showed the complex hysteresis loop due to the effect of slope, residual displacement, and kinematic forces.