• Journal of the Korean Data and Information Science Society
• /
• v.5 no.1
• /
• pp.59-66
• /
• 1994

Heller and Simonoff(1990) compared several methods of estimating the regression coefficient in a modified proportional hazards model, when the response variable is subject to censoring. We give another method of estimating the parameters in the model which also allows the dependent variable to be censored and the error distribution to be unspecified. The proposed method differs from that of Miller(1976) and that of Buckely and James(1979). We also obtain the variance estimator of the coefficient estimator and compare that with the Buckely-James Variance estimator studied by Hillis(1993).

• Proceedings of the KIEE Conference
• /
• 1993.07b
• /
• pp.1038-1040
• /
• 1993

Single phase AC to DC Boost-converter which is control led with a variable hysteretic current mode for improvent of input power factor makes high power factor possible. Power Factor correction circuit can be identified with a determination of each parameters. A simaluation and experiment result to load and parameter variation is examined.

• Journal of the Korea Safety Management & Science
• /
• v.2 no.2
• /
• pp.71-83
• /
• 2000

This paper presents accelerated life tests for Type I censoring data under probabilistic stresses. Probabilistic stress, $S_j$, is the random variable for stress influenced by test environments, test equipments, sampling devices and use conditions. The hazard rate, ,$theta_j$, is the random variable of environments and the function of probabilistic stress. Also it is assumed that the general stress distribution is uniform, the life distribution for the given hazard rate, $\theta$, is exponential and inverse power law model holds. In this paper, we obtained maximum likelihood estimators of model parameters and the mean life in use stress condition.

• Journal of Welding and Joining
• /
• v.28 no.6
• /
• pp.76-83
• /
• 2010

The purpose of this study is to propose a method to decide near-optimal settings of the constant current fuzzy control parameters using a controlled random search. This method tries to find the near-optimal settings of the constant current fuzzy control parameters through experiments. It has an advantage of being able to carry out searches in the search domain which includes some irregular points. The method suggested in this study was used to determine the fuzzy control parameters by which the desired welding current were formed during inverter DC resistance spot welding. The output variable was the ITAE (integral of time multiplied by the absolute error). This output variable was determined according to the input variables, which are the GE, GDE, and GDU. This study described how to obtained near-optimal welding current condition over a wide search space conducting a relatively small number of experiments.

• Structural Engineering and Mechanics
• /
• v.59 no.6
• /
• pp.1037-1054
• /
• 2016

Structural parameter evaluation and external force estimation are two important parts of structural health monitoring. But the structural parameter identification with limited input information is still a challenging problem. A new simultaneous identification method in time domain is proposed in this study to identify the structural parameters and evaluate the external force. Each sampling point in the time history of external force is taken as the unknowns in force evaluation. To reduce the number of unknowns for force evaluation the time domain measurements are divided into several windows. In each time window the structural excitation is decomposed by orthogonal polynomials. The time-variant excitation can be represented approximately by the linear combination of these orthogonal bases. Structural parameters and the coefficients of decomposition are added to the state variable to be identified. The extended Kalman filter (EKF) is augmented and selected as the mathematical tool for the implementation of state variable evaluation. The proposed method is validated numerically with simulation studies of a time-invariant linear structure, a hysteretic nonlinear structure and a time-variant linear shear frame, respectively. Results from the simulation studies indicate that the proposed method is capable of identifying the dynamic load and structural parameters fairly accurately. This method could also identify the time-variant and nonlinear structural parameter even with contaminated incomplete measurement.

• Annals of Clinical Neurophysiology
• /
• v.3 no.1
• /
• pp.31-36
• /
• 2001

Background & Objectives : Fractal Dimension(FD) could be an index of correlation between variable parameters in non-linear chaotic signals. We tried to demonstrate that EEG wave is compatible with chaotic waves by measuring the Lyapunov exponent index and compared the difference of FD between variable age groups(teens, 30's, 50's) Methods : We estimated the Lyapunov exponent index and the FD from digital EEG data among five persons in each normal age groups by using the software which is programmed in our laboratory. Statistical analysis was done with SPSS win 8.0. The statistical differences of Lyapunov exponent index and FD between each electrodes and each age groups were done with ANOVA and paired sample t-test. Result : The Lyapunov exponent indexes were larger than 1 in each electrode and age group. There is no statistical difference in FD between each electrodes and each age groups. Except in 30th age group. In this group the FD of right hemisphere is larger than that of left hemisphere. Conclusion : The result of Lyapunov exponent index means EEG wave is a non-linear chaotic signal. And the results of FD suggest that chaotic parameters of right hemisphere is larger than those of left hemisphere at rest at least in younger people. We think that chaotic parameters can be a useful tool in investigating the variable diseases or brain states.

• The Transactions of the Korean Institute of Power Electronics
• /
• v.14 no.4
• /
• pp.293-298
• /
• 2009

This paper presents a variable gain current control algorithm for the stabilized grid connection between the grid and a doubly fed induction(DFIG) as a wind power generator. The performance of a RSC current controller depends highly on accurate machine parameters, and especially requires a fast and robust response regardless of the disturbances such as voltage sag. However, parameter variations of a DFIG occur at the point of grid connection, which affects the current controller gains based on DFIG parameters after a DFIG is connected to the grid. Thus, performance degrades when actual machine parameters depart from values used in the control system. In the proposed algorithm, current controller gains of the rotor side converter(RSC) are changed after a DFIG is connected to the grid. The simulation results and experimental results for a 750kW are shown to illustrate the feasibility of the proposed algorithm.

• Journal of Korean Neurosurgical Society
• /
• v.47 no.2
• /
• pp.128-133
• /
• 2010

Objective : The secondary verification of Leksell Gamma Knife treatment planning system (LGP) (which is the primary verification system) is extremely important in order to minimize the risk of treatment errors. Although prior methods have been developed to verify maximum dose and treatment time, none have studied maximum dose coordinates and treatment volume. Methods : We simulated the skull shape as an ellipsoid with its center at the junction between the mammillary bodies and the brain stem. The radiation depths of the beamlets emitted from 201 collimators were calculated based on the relationship between this ellipsoid and a single beamlet expressed as a straight line. A computer program was coded to execute the algorithm. A database system was adopted to log the doses for $31{\times}31{\times}31$ or 29,791 matrix points allowing for future queries to be made of the matrix of interest. Results : When we compared the parameters in seven patients, all parameters showed good correlation. The number of matrix points with a dose higher than 30% of the maximal dose was within ${\pm}\;2%$ of LGP. The 50% dose volume, which is generally the target volume, differs maximally by 4.2%. The difference of the maximal dose ranges from 0.7% to 7%. Conclusion : Based on the results, the variable ellipsoid modeling technique or variable ellipsoid modeling technique (VEMT) can be a useful and independent tool to verify the important parameters of LGP and make up for LGP.

• Geomechanics and Engineering
• /
• v.21 no.4
• /
• pp.357-369
• /
• 2020

Different from the conventional planar fracture and simplified Newton model, for power-law slurries with a lower water-cement ratio commonly used in grouting engineering, flow model in geological rough fractures is built based on ten standard profiles from Barton (1977) in this study. The numerical algorithm is validated by experimental results. The flow mechanism, grout superiority, and water plugging of pseudo plastic slurry are revealed. The representations of hydraulic grouting properties for JRCs are obtained. The results show that effective plugging is based on the mechanical mechanisms of the fluctuant structural surface and higher viscosity at the middle of the fissure. The formulas of grouting parameters are always variable with the roughness and shear movement, which play a key role in grouting. The roughness can only be neglected after reaching a threshold. Grouting pressure increases with increasing roughness and has variable responses for different apertures within standard profiles. The whole process can be divided into three stationary zones and three transition zones, and there is a mutation region (10 < JRCs < 14) in smaller geological fractures. The fitting equations of different JRCs are obtained of power-law models satisfying the condition of -2 < coefficient < 0. The effects of small apertures and moderate to larger roughness (JRCs > 10.8) on the permeability of surfaces cannot be underestimated. The determination of grouting parameters depends on the slurry groutability in terms of its weakest link with discontinuous streamlines. For grouting water plugging, the water-cement ratio, grouting pressure and grouting additives should be determined by combining the flow conditions and the apparent widths of the main fracture and rough surface. This study provides a calculation method of grouting parameters for variable cement-based slurries. And the findings can help for better understanding of fluid flow and diffusion in geological fractures.

• Proceedings of the KIEE Conference
• /
• 1998.07a
• /
• pp.146-148
• /
• 1998

For the shape design of electromagnetic devices using the FEM, the choice of design parameters influence to the success of the optimization process. If the design parameter distribution has a one to one corespondence with finite element model, we can encounter not only serious accuracy problem but also obtain a zigzag shape along the interface. The nodes between those design parameters can be parameterized by interpolating using one among many interpolation methods. The conventional parameterization of design parameters has a limit of application for shape, because design parameters and movable nodes are linearly intepolated. In this paper, using the B-spline curve that use to present any interfaces in computer graphics, the curvilinear parameterization between design parameters and node points is compared with the linear parameterization.