• Geophysics and Geophysical Exploration
• /
• v.23 no.3
• /
• pp.157-167
• /
• 2020

With the recent advancement of computer hardware and the contribution of open source libraries to facilitate access to artificial intelligence technology, the use of machine learning (ML) and deep learning (DL) technologies in various fields of exploration geophysics has increased. In addition, ML researchers have developed complex algorithms to improve the inference accuracy of various tasks such as image, video, voice, and natural language processing, and now they are expanding their interests into the field of automatic machine learning (AutoML). AutoML can be divided into three areas: feature engineering, architecture search, and hyperparameter search. Among them, this paper focuses on hyperparamter search with Bayesian optimization, and applies it to the problem of facies classification using seismic data and well logs. The effectiveness of the Bayesian optimization technique has been demonstrated using Vincent field data by comparing with the results of the random search technique.

• 한국방재학회:학술대회논문집
• /
• 2009.02b
• /
• pp.43-50
• /
• 2009

Strong ground motion attenuation relationship represents a comprehensive trend of ground shakings at sites with distances from the source, geology, local soil conditions, and others. It is necessary to develop an attenuation relationship with careful considerations of characteristics of the target area for reliable seismic hazard/risk assessments. In the study, observed ground motions from the January 2007 magnitude 4.9 Odaesan earthquake and the events occurring in the Gyeongsang provinces are compared with the previously proposed ground attenuation relationships in the Korean Peninsula to select most appropriate one. In the meantime, a few strong ground motion attenuation relationships are proposed and introduced in HAZUS, which have been designed for the Western United States and the Central and Eastern United States. The selected relationship from the ones for the Korean Peninsula has been compared with attenuation relationships available in HAZUS. Then, the attenuation relation for the Western United States proposed by Sadigh et al. (1997) for the Site Class B has been selected for this study. Reliability of the assessment will be improved by using an appropriate attenuation relation. It has been used for the earthquake loss estimation of the Gyeongju area located in southeast Korea using the deterministic method in HAZUS with a scenario earthquake (M=6.7). Our preliminary estimates show 15.6% damage of houses, shelter needs for about three thousands residents, and 75 life losses in the study area for the scenario events occurring at 2 A.M. Approximately 96% of hospitals will be in normal operation in 24 hours from the proposed event. Losses related to houses will be more than 114 million US dollars. Application of the improved methodology for loss estimation in Korea will help decision makers for planning disaster responses and hazard mitigation.

• Journal of the Earthquake Engineering Society of Korea
• /
• v.6 no.5
• /
• pp.1-17
• /
• 2002

Prognostic characteristics of hypothetical tsunamis in the East Sea are further discussed based on numerical simulations using linear long wave theory than the last paper(Choi et al). As for choice of source zones, we used 28 cases based on fault parameters of hypothetical earthquakes and 76 cases based on simple initial surface shapes of tsunamigenic earthquakes selected by the seismic gap theory. As a result, the wave heights along the whole coastline adjacent to the East See of tsunamis due to these hypothetical earthquake are presented. Analyses also lead us to conclude that the selection of geographical zones with low risk of tsunamis can be used as a tool for coastal disaster mitigation planning.

• Structural Engineering and Mechanics
• /
• v.9 no.3
• /
• pp.241-256
• /
• 2000

The major sources of energy dissipation in steel frames with partially restrained (PR) connections are evaluated. Available experimental results are used to verify the mathematical model used in this study. The verified model is then used to quantify the energy dissipation in PR connections due to hysteretic behavior, due to viscous damping and at plastic hinges if they are formed. Observations are made for two load conditions: a sinusoidal load applied at the top of the frame, and a sinusoidal ground acceleration applied at the base of the frame representing a seismic loading condition. This analytical study confirms the general behavior, observed during experimental investigations, that PR connections reduce the overall stiffness of frames, but add a major source of energy dissipation. As the connections become stiffer, the contribution of PR connections in dissipating energy becomes less significant. A connection with a T ratio (representing its stiffness) of at least 0.9 should not be considered as fully restrained as is commonly assumed, since the energy dissipation characteristics are different. The flexibility of PR connections alters the fundamental frequency of the frame. Depending on the situation, it may bring the frame closer to or further from the resonance condition. If the frame approaches the resonance condition, the effect of damping is expected to be very important. However, if the frame moves away from the resonance condition, the energy dissipation at the PR connections is expected to be significant with an increase in the deformation of the frame, particularly for low damping values. For low damping values, the dissipation of energy at plastic hinges is comparable to that due to viscous damping, and increases as the frame approaches failure. For the range of parameters considered in this study, the energy dissipations at the PR connections and at the plastic hinges are of the same order of magnitude. The study quantitatively confirms the general observations made in experimental investigations for steel frames with PR connections; however, proper consideration of the stiffness of PR connections and other dynamic properties is essential in predicting the dynamic behavior.

• The Journal of the Acoustical Society of Korea
• /
• v.11 no.3
• /
• pp.53-60
• /
• 1992

Many adaptive beamforming methods have been studied for interference cancellation and speech signal enhancement in telephone conference and auditorium. Main aspect of adaptive beamforming methods for speech signal processing is different from radar, sonar and seismic signal processing because desire output signal should be apt to the human ear. Considering that phase of speech is quite insensible to the human ear, Sondhi proposed a nonlinear constrained optimization technique whose constraint was on the magnitude transfer function from the source to the output. In real environment the phase response of the speech signal affects the human auditorium system. So it is desirable to design linear phase system. In this paper, linear phase beamformer is proposed and sample processing algorithm is also proposed for real time consideration Simulation results show that the proposed algorithm yields more consistent beam patterns and deep nulls to the noise direction than Sondhi's.

• Smart Structures and Systems
• /
• v.6 no.3
• /
• pp.197-209
• /
• 2010

The structural state of a bridge is currently examined by visual inspection or by wired sensor techniques, which are relatively expensive, vulnerable to inclement conditions, and time consuming to undertake. In contrast, wireless sensor networks are easy to deploy and flexible in application so that the network can adjust to the individual structure. Different sensing techniques have been used with such networks, but the acoustic emission technique has rarely been utilized. With the use of acoustic emission (AE) techniques it is possible to detect internal structural damage, from cracks propagating during the routine use of a structure, e.g. breakage of prestressing wires. To date, AE data analysis techniques are not appropriate for the requirements of a wireless network due to the very exact time synchronization needed between multiple sensors, and power consumption issues. To unleash the power of the acoustic emission technique on large, extended structures, recording and local analysis techniques need better algorithms to handle and reduce the immense amount of data generated. Preliminary results from utilizing a new concept called Acoustic Emission Array Processing to locally reduce data to information are presented. Results show that the azimuthal location of a seismic source can be successfully identified, using an array of six to eight poor-quality AE sensors arranged in a circular array approximately 200 mm in diameter. AE beamforming only requires very fine time synchronization of the sensors within a single array, relative timing between sensors of $1{\mu}s$ can easily be performed by a single Mote servicing the array. The method concentrates the essence of six to eight extended waveforms into a single value to be sent through the wireless network, resulting in power savings by avoiding extended radio transmission.

• Geophysics and Geophysical Exploration
• /
• v.1 no.3
• /
• pp.161-169
• /
• 1998

The Kunsan basin is a pull-apart basin which was formed during Tertiary. The pre-Tertiary section consists of various rock types, such as meta-sediments, igneous rocks, carbonates, clastics, and volcanics. Tertiary sections are the main targets for the petroleum exploration. In order to determine the optimum processing parameters of the basin, about 12 kinds of test processings were performed. The first main steps for the quality control is to determine the noisy or bad traces by examining the near trace section and shot gathers. The true amplitude recovery was applied to account for the amplitude losses due to spherical divergence and inelastic attenuation. Source designature and predictive deconvolution test were conducted to determine the optimum wavelet parameters and to remove the multiples. Velocity analysis was performed at 1km intervals. The optimum mute function was picked by locating the range of offsets which gives the best stacking response for any particular reflections. Post-stack deconvolution was tested to see if the quality of stacked data improved. The stacked data was migrated using a finite difference algorithm. The migration velocity was obtained from the stacking velocities using the time varying percentages. The AGC sections were provided for the structural interpretation. The RAP sections were used for DHI analysis and for the detection of volcanics.

• Geophysics and Geophysical Exploration
• /
• v.3 no.2
• /
• pp.70-75
• /
• 2000

In the seismic migration, Kirchhoff and reverse time migration are used in general. In the reverse time migration using wave equation, two-way and one-way wave equation are applied. The approach of one-way wave equation uses approximately computed downward continuation extrapolator, it need tess amounts of calculations and core memory in compared to that of two-way wave equation. In this paper, we applied one-way wave equation to pre-stack reverse time migration. In the frequency-space domain, forward propagation of source wavefield and back propagration of measured wavefield were executed by using monochromatic one-way wave equation, and zero-lag cross correlation of two wavefield resulted in the image of subsurface. We had implemented prestack migration on a massively parallel processors (MPP) CRAYT3E, and knew the algorithm studied here is efficiently applied to the prestck migration due to its suitability for parallelization.

• Proceedings of the Korean Geotechical Society Conference
• /
• 2008.10a
• /
• pp.1415-1426
• /
• 2008

Stone columns, locally called "GCP (granular compaction pile)" can be used to improve strength and resistance against lateral movement of a foundation soil like rigid piles and piers. Also installation of such a discrete column facilitates drainage, and densifies and reinforces the soil in the sense of ground improvement. The integrity of the GCP has been indirectly controlled with the records of each batch including depth and the quantity of stone filled. An integrity testing was attempted using crosshole S-wave logging. The method is conceptionally same as the crosshole sonic logging (CSL) for drilled piers. The only and critical difference is that S-wave should be used in the logging, because P-wave velocity of the stone column is less than that of ground water. The crosshole sonic logger does not have the capability to measure S-wave propagating through the skeleton of crushed stone. An electro-mechanical source, which can generate either P- or SH-waves, and a 1-D geophone were used to measure SH-waves. Two 76mm diameter cased boreholes were installed 1 meter apart across the nominal 700mm diameter stone column. At every 10cm of depth, shear wave was measured across the stone column. One more borehole was also installed 1 meter outward from the one of the above boreholes to measure the shear wave profile of the surrounding soil. The diametric variation of the stone column with respect to depth was evaluated from the shear wave arrival times across the stone column, and shear wave velocities of crushed stone and surrounding soil. The volume calculated with these variational diameters is very close to the actual quantity of the stone filled.

• International Journal of Highway Engineering
• /
• v.12 no.4
• /
• pp.139-145
• /
• 2010

An integrity testing for stone columns was attempted using crosshole S-wave logging. The method is conceptionally quite similar to the crosshole sonic logging (CSL) for drilled piers. The critical difference in the logging is the use of S-wave rather than P-wave, which is used in CSL, because swave is the only wave sensing the stiffness of slower unbounded materials than water. An electro-mechanical source, which can generate reversed Swave signals, was utilized in the logging. The stone column was delineated using the S-wave travel times across the stone column, the S-wave velocity profile of the crushed stone($V_{cs}$-profile) and that of surrounding soil($V_s$-profile). In the calculation of $V_{cs}$-profile of the crushed stone, its friction angle and Ko (coefficient of lateral earth pressure at rest) are recommended to be used. The calculation of the column diameter is not much affected by the values of friction angle and Ko.