• Journal of the Korean Institute of Telematics and Electronics B
• /
• v.31B no.4
• /
• pp.145-154
• /
• 1994

This paper proposes an efficient method for improving the training performance of the neural network by using a hybrid of a stochastic approximation and a backpropagation algorithm. The proposed method improves the performance of the training by appliying a global optimization method which is a hybrid of a stochastic approximation and a backpropagation algorithm. The approximate initial point for a stochastic approximation and a backpropagation algorihtm. The approximate initial point for fast global optimization is estimated first by applying the stochastic approximation, and then the backpropagation algorithm, which is the fast gradient descent method, is applied for a high speed global optimization. And further speed-up of training is made possible by adjusting the training parameters of each of the output and the hidden layer adaptively to the standard deviation of the neuron output of each layer. The proposed method has been applied to the parity checking and the pattern classification, and the simulation results show that the performance of the proposed method is superior to that of the backpropagation, the Baba's MROM, and the Sun's method with randomized initial point settings. The results of adaptive adjusting of the training parameters show that the proposed method further improves the convergence speed about 20% in training.

• Structural Engineering and Mechanics
• /
• v.56 no.6
• /
• pp.959-982
• /
• 2015

In this study, a non-stationary random earthquake Clough-Penzien model is used to describe earthquake ground motion. Using stochastic direct integration in combination with an equivalent linear method, a solution is established to describe the non-stationary response of lead-rubber bearing (LRB) system to a stochastic earthquake. Two parameters are used to develop an optimization method for bearing design: the post-yielding stiffness and the normalized yield strength of the isolation bearing. Using the minimization of the maximum energy response level of the upper structure subjected to an earthquake as an objective function, and with the constraints that the bearing failure probability is no more than 5% and the second shape factor of the bearing is less than 5, a calculation method for the two optimal design parameters is presented. In this optimization process, the radial basis function (RBF) response surface was applied, instead of the implicit objective function and constraints, and a sequential quadratic programming (SQP) algorithm was used to solve the optimization problems. By considering the uncertainties of the structural parameters and seismic ground motion input parameters for the optimization of the bearing design, convex set models (such as the interval model and ellipsoidal model) are used to describe the uncertainty parameters. Subsequently, the optimal bearing design parameters were expanded at their median values into first-order Taylor series expansions, and then, the Lagrange multipliers method was used to determine the upper and lower boundaries of the parameters. Moreover, using a calculation example, the impacts of site soil parameters, such as input peak ground acceleration, bearing diameter and rubber shore hardness on the optimization parameters, are investigated.

• Structural Engineering and Mechanics
• /
• v.3 no.4
• /
• pp.313-324
• /
• 1995

Flexible cantilever pipes conveying fluids with high velocity are analysed for their dynamic response and stability behaviour. The Young's modulus and mass per unit length of the pipe material have a stochastic distribution. The stochastic fields, that model the fluctuations of Young's modulus and mass density are characterized through their respective means, variances and autocorrelation functions or their equivalent power spectral density functions. The stochastic non self-adjoint partial differential equation is solved for the moments of characteristic values, by treating the point fluctuations to be stochastic perturbations. The second-order statistics of vibration frequencies and mode shapes are obtained. The critical flow velocity is first evaluated using the averaged eigenvalue equation. Through the eigenvalue equation, the statistics of vibration frequencies are transformed to yield critical flow velocity statistics. Expressions for the bounds of eigenvalues are obtained, which in turn yield the corresponding bounds for critical flow velocities.

• Structural Engineering and Mechanics
• /
• v.39 no.4
• /
• pp.541-558
• /
• 2011

This paper presents an analysis of stochastic eigenvalue problem of plane bar structures. Particular attention is paid to the effect of spatial variations of the flexural properties of the structure on the first four eigenvalues. The problem of spatial variations of the structure properties and their effect on the first four eigenvalues is analyzed in detail. The stochastic eigenvalue problem was solved independently by stochastic finite element method (stochastic FEM) and Monte Carlo techniques. It was revealed that the spatial variations of the structural parameters along the structure may substantially affect the eigenvalues with quite wide gap between the two extreme cases of zero- and full-correlation. This is particularly evident for the multi-segment structures for which technology may dictate natural bounds of zero- and full-correlation cases.

• Advances in materials Research
• /
• v.13 no.4
• /
• pp.299-319
• /
• 2024

Due to higher strength-to-weight ratio of composite laminates, they find uses in many weight-sensitive applications like aerospace, automobile and marine structures. From a reliability point of view, accurate prediction of failure of these structures is important. Due to the complexities in the manufacturing processes of composite laminates, there is a variation in the material properties and geometric parameters. Hence stochastic aspects are important while designing the composite laminates. Many existing works of composite laminate failure analysis are based on the deterministic approach but it is important to consider the randomness in the material properties, geometry and loading to predict accurate failure loads. In this paper the statistics of the ultimate failure load of the [0/θ]_s laminated composite plate (LCP) containing the edge crack and voids subjected to the tensile loading are presented in terms of the mean and coefficient of variance (COV). The objective is to better the efficacy of laminate failure by predicting the statistics of the ultimate failure load of LCP with random material, geometric and loading parameters. The stochastic analysis is done by using the extended finite element method (XFEM) combined with the second-order perturbation technique (SOPT). The ultimate failure load of the LCP is obtained by ply-by-ply failure analysis using the ply discount method combined with the Tsai-Wu failure criterion. The aim is to know the effect of the stacking sequence, crack length, crack angle, location of voids and number of voids on the mean and corresponding COV of the ultimate failure load of LCP is investigated. The results of the ultimate failure load obtained by the present method are in good agreement with the existing experimental and numerical results. It is observed that [0/θ]_s LCPs are very sensitive to the randomness in the crack length, applied load, transverse tensile strength of the laminate and modulus of elasticity of the material, so precise control of these parameters is important. The novelty of the present study is, the stochastic implementation in XFEM for the failure prediction of LCPs containing crack and voids.

• Bulletin of the Korean Mathematical Society
• /
• v.56 no.2
• /
• pp.479-489
• /
• 2019

In this paper, we show the existence of a weak solution for a stochastic differential equation driven by an additive fractional Brownian sheet with Hurst parameters H, H' > 1/2, and a drift coefficient satisfying the linear growth condition. The result is obtained using a suitable Girsanov theorem for the fractional Brownian sheet.

• Proceedings of the Earthquake Engineering Society of Korea Conference
• /
• 2000.04a
• /
• pp.70-80
• /
• 2000

In order to reduce seismic hazard the characteristics of strong earthquakes are required. In the region where strong earthquakes do not happen frequently the stochastic simulation of strong motion is an alternative way to predict strong motions. this simulation required input parameters such as the quality factor the corner frequency the moment magnitude the stress drop and so on which can be obtained from analyses of records of small and intermediate earthquakes. Using those parameters obtained in the previous work the strong ground motions are predicted employing the stochastic method, . The results are compared to the two observed earthquakes-the Ulsan Offshore Earthquake and the Kyungju Earthquake. Although some deviations are found the predictions are similar to the observed data. Finally we computed attenuation equations for PGA PGV and ground accelerations for some frequencies using the results of predictions. These results can be used for earthquake engineering and more reliable results will come out as earthquake observations continue.

• 제어로봇시스템학회:학술대회논문집
• /
• 1991.10b
• /
• pp.1858-1863
• /
• 1991

The theory of stock option pricing has, recently, attracted attention of many researchers interested not only in finance but also in statistics and control theory. In this field, the problem of estimating stock return volatility is, above all, of great importance in calculating actual stock option value. In this paper, we assume that the stock market is represented by the stochastic volatility model which is the same as that of Hull and White. Then, we propose an approximation function of option value. It is a type of Black-Sholes option formula in which the first and the second order moments of logarithmic stock value are modified in a special form from the original model. Finally, an algorithm of estimating the parameters of the stochastic volatility model is given, and parameters are estimated by using Nikkei 225 index option data.

• The Transactions of The Korean Institute of Electrical Engineers
• /
• v.57 no.1
• /
• pp.20-24
• /
• 2008

A bid-based pool(BBP) model is representative of energy market structure in a number of restructured electricity markets. Supply function equilibrium(SFE) models of interaction better match what is explicitly required in the bid formats of typical BBP markets. Many of the results in the SFE literature involve restrictive parametrization of the bid cost functions. In the SFE models, two parameters, intercept and slope, are available for strategic bidding. This paper addresses the realistic competition format that players can choose both parameters arbitrarily. In a fixed demand function, equilibrium conditions for generation company's profit maximization have a degree of freedom, which induces multi-equilibrium. So it is hard to choose a convergent equilibrium. However, consideration of stochastic demand function makes the equilibrium conditions independent each other based on the amount of variance of stochastic demand function. This variance provides the bidding players with incentives to change the slope parameter from an equilibrium for a fixed demand function until the slope parameter equilibrium.

• Coupled systems mechanics
• /
• v.7 no.2
• /
• pp.211-232
• /
• 2018

In this work we present an upscaling technique for multi-scale computations based on a stochastic model calibration technique. We consider a coarse-scale continuum material model described in the framework of generalized standard materials. The model parameters are considered uncertain, and are determined in a Bayesian framework for the given fine scale data in a form of stored energy and dissipation potential. The proposed stochastic upscaling approach is independent w.r.t. the choice of models on coarse and fine scales. Simple numerical examples are shown to demonstrate the ability of the proposed approach to calibrate coarse scale elastic and inelastic material parameters.