• Proceedings of the Korean Society of Computer Information Conference
• /
• 2021.07a
• /
• pp.75-77
• /
• 2021

전력 생산의 효율을 높이고 지속적인 공정관리를 위해 전력 설비 데이터의 패턴을 분석하고 원인이 되는 주요 변수를 찾는 것이 중요하다. 따라서, 본 연구에서는 전력 설비 데이터의 패턴을 분석하기 위해 데이터를 군집화하고 연구 방법으로 Decision Tree, Random Forest와 ResNet을 이용하여 패턴을 분류하였다. Class Activation Map을 이용하여 설비데이터의 원인이 되는 주요 변수를 확인하였다. 본 연구를 통해 전력 설비 데이터의 분류 및 원인 분석이 가능한 통합적 솔루션을 제시하고자 한다.

• The Journal of the Korea institute of electronic communication sciences
• /
• v.16 no.2
• /
• pp.287-292
• /
• 2021

In this paper, we propose an efficient CNN application method by analyzing the effect of activation function on the failure diagnosis of the inductive motor stator. Generally, the main purpose of the inductive motor stator failure diagnosis is to prevent the failure by rapidly diagnosing the minute turn short. In the application of activation function, experiments show that the Sigmoid function is 23.23% more useful in accuracy of diagnosis than the ReLu function, although it is shown that ReLu has superiority in overall fixer failure in utilizing the activation function.

• Journal of the Korean Institute of Gas
• /
• v.4 no.4 s.12
• /
• pp.42-49
• /
• 2000

Due to their size and complexity, it is very difficult to make diagnostic system for the whole chemical processes. Therefore, a systematic approach is required to decompose larpge-scale process into sub-processes and then diagnose them. This paper suggests a method for the minimization of knowledge base and flexible diagnosis to be used in qualitative fault diagnosis based on Fault-Effect Tree model. The system can be decomposed for flexible diagnosis, size reduction of knowledge base, and consistent construction of complex knowledge base. The new node, gate-variable, is introduced to connect the cause-effect relationships of each sub-process. For on-line diagnosis, off-line analysis is performed to construct Fault-Effect Trees of gate-variables as well as activation conditions of gate-variables. On-line diagnosis strategy is modified to get the same diagnosis result without system decomposition. The proposed method is illustrated with a fault diagnosis system for a large-scale boiler plant.

• Proceedings of the KIEE Conference
• /
• 2011.07a
• /
• pp.93-94
• /
• 2011

This paper analyzes the shaft torsion of a doubly-fed induction generator (DFIG) for a wind farm collector system fault. When a fault occurs, the active power of the DFIG cannot be transmitted to the grid and thus accelerates the rotation of both the blade and the rotor. Due to the different inertia of these, the angle of deviation fluctuates and the shaft torsion is occurred. This becomes much severe when the rotational speed of the blade exceeds a threshold, which activating the pitch control to reduce the mechanical power. The torque, which can be sixty times larger than that in the steady state, may destroy the shaft. The shaft torsion phenomena are simulated using the EMTP-RV simulator. The results indicate that when a wind farm collector system fault occurs, a severe shaft torsion is occurred due to the activation of the pitch control.

• Economic and Environmental Geology
• /
• v.50 no.2
• /
• pp.97-104
• /
• 2017

Here we present the timings of reactivated events from a fault in the northern Pohang area, which should be located at the northern-end of Yangsan fault line, the major fault in the southeastern Korean Peninsula. Recently developed illite-age-analysis (IAA) approach was employed for determining the fault-activated timing, combined with illite-polytype quantification using the optimized full-pattern-fitting (FPF) method, and K-Ar age-dating for each size fraction($<0.1{\mu}m$, $0.1-0.4{\mu}m$, and $0.4-1.0{\mu}m$) of 4 fault clay samples. Two chronological records of brittle fault-activation events were recognized at $19.6{\pm}1.86Ma$ and $26.1{\pm}2.55-27.9{\pm}3.46Ma$. The ages are much younger than those of fault clays from Sangcheon-ri area (41.5~43.5 and 50.7 Ma), the southern part of Yangsan fault line, and are close to the timing of East Sea-opening event. Further chronological analysis for additional sites of the Yangsan fault should be needed to reveal the time-scheme of the tectonic events and their spatial distributions along the fault line.

• Economic and Environmental Geology
• /
• v.56 no.3
• /
• pp.293-300
• /
• 2023

The Shape Preferred Orientation (SPO) method has been used to analyze the orientation of fault motion, which is utilized as basic data for fault kinematics studies. The rigid grains, which as quartz, feldspar, and rock fragments, in the fault gouge are arranged in the P-shear direction through rigid body rotation by a given shear stress. Using this characteristic, the fault motion can be estimated from the SPO inversely. Recently, a method for securing precision and reliability by measuring 3D-SPO using X-ray CT images and examining the shape of a large number of particles in a short time has been developed. As a result, the SPO method analyzes the orientation of thousands to tens of thousands of particles at high speed, suggests the direction of fault motion, and provides easy accessibility and reliable data. In addition, the shape information and orientation distribution data of particles, which are by-products obtained in the SPO analysis process, are expected to be used as basic data for conducting various studies such as the local deformation of fault rocks and the fault generation mechanism.

• Journal of the Korean Society of Industry Convergence
• /
• v.7 no.4
• /
• pp.349-354
• /
• 2004

The growth mode of the Si layers which were grown on Si(111) by using Ag as surfactant were investigated by intensity oscillations of the RHEED specular spot at the different temperatures. we found that the introduction of Ag as the surfactant alters the growth mode from a three-dimensional clustering mechanism to a two-dimensional layer-by-layer growth. In the growth of Si layers on Si(111) with a surfactant Ag, At $450^{\circ}C$, RHEED intensity oscillation was very stable and periodic from early stage of deposition to 32 ML. RHEED patterns during homoepitaxial growth at $450^{\circ}C$ was changed from $7{\times}7$ structure into ${\sqrt{3}}{\times}{\sqrt{3}}$ structures. Since the ${\sqrt{3}}{\times}{\sqrt{3}}$ structure include no stacking fault, the stacking fault layer seems to be reconstructed into normal stacking one at transition from the $7{\times}7$ structure to a ${\sqrt{3}}{\times}{\sqrt{3}}$ one. We also found that the number of the intensity oscillation of the specular spot for Si growth with a surfactant Ag was more than for Si growth without a surfactant. This result may be explained that the activation energy decrease for the surface diffusion of Si atoms due to segregation of the surfactant toward the growing surface.

• Economic and Environmental Geology
• /
• v.33 no.1
• /
• pp.61-75
• /
• 2000

Through modeling fault network using thin plate finite element technique in the San Andreas Fault system with slip rate over 1mm/year, as well as elevation, heat flow, earthquakes, geodetic data and crustal thickness, we compare the results with velocity boundary conditions of plate based on the NUVEL-1 plate model and the approximation of deformation in the Great Basin region. The frictional and dislocation creep constants of the crust are calculated to reproduce the observed variations in the maximum depth of seismicity which corresponds to the temperature ranging from $350^{\circ}C$ to $410^{\circ}C$. The rheologic constants are defined by the coefficient of friction on faults, and the apparent activation energy for creep in the lower crust. Two parameters above represent systematic variations in three experiments. The pattern of model indicates that the friction coefficient of major faults is 0.17~0.25. we test whether the weakness of faults is uniform or proportional to net slip. The geologic data show a good agreement when fault weakness is a trend of an additional 30% slip dependent weakening of the San Andreas. The results of study suggest that all weakening is slip dependent. The best models can be explained by the available data with RMS mismatch of as little as 3mm/year, so their predictions can be closely related with seismic hazard estimation, at least along faults where no data are available.

• Journal of the Korean Institute of Illuminating and Electrical Installation Engineers
• /
• v.26 no.7
• /
• pp.70-76
• /
• 2012

In general, the unidirectional power flow is normal in distribution feeders before activation of distributed power source such as PV. However, the interactive power flow is likely to occur in case of the power system under distributed generation. This interactive power flow can cause an unexpected effect on convectional protection coordination systems designed based only on the unidirectional power flow system. When the power line system encounters a problem, the interactive power flow can be a contributed current source and this makes the fault current bigger or smaller compared to the unidirectional case. The effect of interactive power flow is varied depending on the location of the point to ground fault, relative location of the PV, and connection method. Therefore it is important to analyse characteristics of fault current and interactive flow for various transformer connection and location of the PV. This paper proposes a method of improved protection coordination which can be adopted in the protective device for customers in distribution feeders interconnected with the PV. The proposed method is simulated and analysed using PSCAD/EMTDC under various conditions.

• Proceedings of the KIEE Conference
• /
• 2011.07a
• /
• pp.706-707
• /
• 2011

The insertion of the crowbar system in the doubly fed induction generator rotor circuit for a short period of time during grid disturbance enables a more efficient way of limiting transient rotor current and hence protecting the rotor side converter (RSC) and the DC - link capacitor. When crowbar is activated at fault occurrence and clearance time, RSC is blocked while DC -link capacitor and the grid side converter (GSC) can be controlled to provide reactive power support at the PCC and improve the voltage which helps to comply with grid codes. In this paper, control strategies for crowbar system to limit the rotor current during fault is presented with RSC and GSC controllers are modified to control PCC voltage during disturbance to enhance DFIG wind farm to comply with some strict grid codes. Model simulated on MATLAB/Simulink verify the study through simulation results presented.