• Journal of the Korean Society of Industry Convergence
• /
• v.3 no.4
• /
• pp.361-367
• /
• 2000

This paper proposes an efficient feature extraction of the image data using nonlinear principal component analysis neural networks of a new learning algorithm. The proposed method is a learning algorithm with momentum for reflecting the past trends. It is to get the better performance by restraining an oscillation due to converge the global optimum. The proposed algorithm has been applied to the cancer image of $256{\times}256$ pixels and the coin image of $128{\times}128$ pixels respectively. The simulation results show that the proposed algorithm has better performances of the convergence and the nonlinear feature extraction, in comparison with those using the backpropagation and the conventional nonlinear PCA neural networks.

• Bulletin of the Korean Mathematical Society
• /
• v.49 no.4
• /
• pp.705-714
• /
• 2012

In this paper, we study nonlinear equation arising in MEMS modeling electrostatic actuation. We will prove the local and global existence of solutions of the generalized parabolic MEMS equation. We present that there exists a constant ${\lambda}^*$ such that the associated stationary problem has a solution for any ${\lambda}$ < ${\lambda}^*$ and no solution for any ${\lambda}$ > ${\lambda}^*$. We show that when ${\lambda}$ < ${\lambda}^*$ the global solution converges to its unique maximal steady-state as $t{\rightarrow}{\infty}$. We also obtain the condition for the existence of a touchdown time $T{\leq}{\infty}$ for the dynamical solution. Furthermore, there exists $p_0$ > 1, as a function of $p$, the pull-in voltage ${\lambda}^*(p)$ is strictly decreasing with respect to 1 < $p$ < $p_0$, and increasing with respect to $p$ > $p_0$.

• Steel and Composite Structures
• /
• v.43 no.2
• /
• pp.153-164
• /
• 2022

In this study, a series of cyclic loading tests were performed on viscoelastic dampers (VED) composed of viscoelastic polymer composite material and thin steel plates to observe the variation of the mechanical properties under different loading conditions. A mathematical model was developed based on the Kelvin-Voigt and Bouc-Wen models to formulate the nonlinear force-displacement relationship of the viscoelastic damper. The accuracy of the proposed mathematical model was verified using the data obtained from the tests. The mathematical model was applied to analyze a reinforced concrete framed structure retrofitted with viscoelastic dampers. Nonlinear dynamic analysis results showed that the average maximum inter-story drift ratios of the retrofitted structure met the target limit state after installing the VED. In addition, both the maximum and residual displacements were significantly reduced after the installation of the VED.

• ETRI Journal
• /
• v.45 no.1
• /
• pp.119-130
• /
• 2023

We propose a quantized gradient search algorithm that can achieve global optimization by monotonically reducing the quantization step with respect to time when quantization is composed of integer or fixed-point fractional values applied to an optimization algorithm. According to the white noise hypothesis states, a quantization step is sufficiently small and the quantization is well defined, the round-off error caused by quantization can be regarded as a random variable with identically independent distribution. Thus, we rewrite the searching equation based on a gradient descent as a stochastic differential equation and obtain the monotonically decreasing rate of the quantization step, enabling the global optimization by stochastic analysis for deriving an objective function. Consequently, when the search equation is quantized by a monotonically decreasing quantization step, which suitably reduces the round-off error, we can derive the searching algorithm evolving from an optimization algorithm. Numerical simulations indicate that due to the property of quantization-based global optimization, the proposed algorithm shows better optimization performance on a search space to each iteration than the conventional algorithm with a higher success rate and fewer iterations.

• Journal of Advanced Marine Engineering and Technology
• /
• v.20 no.2
• /
• pp.101-101
• /
• 1996

This paper presents the linearized fuzzy modeling technique of nonlinear dynamical system and the stability analysis of fuzzy control system. Firstly, the nonlinear system is partitionized by multiple linear fuzzy subcontrol systems based on fuzzy linguistic variables and fuzzy rules. Secondly, the disturbance adaptaion controllers which guarantee the global asymptotic stability of each fuzzy subsystem by an optimal feedback control law are designed and the stability analysis procedures of the total fuzzy control system using Lyapunov functions and eigenvalues are discussed in detail through a given illustrative example.

• Journal of Advanced Marine Engineering and Technology
• /
• v.20 no.2
• /
• pp.33-39
• /
• 1996

This paper presents the linearized fuzzy modeling technique of nonlinear dynamical system and the stability analysis of fuzzy control system. Firstly, the nonlinear system is partitionized by multiple linear fuzzy subcontrol systems based on fuzzy linguistic variables and fuzzy rules. Secondly, the disturbance adaptaion controllers which guarantee the global asymptotic stability of each fuzzy subsystem by an optimal feedback control law are designed and the stability analysis procedures of the total fuzzy control system using Lyapunov functions and eigenvalues are discussed in detail through a given illustrative example.

• Proceedings of the Computational Structural Engineering Institute Conference
• /
• 2003.10a
• /
• pp.145-152
• /
• 2003

An optimal design method in cooperated with nonlinear inelastic analysis method is presented. The proposed nonlinear inelastic method overcomes the difficulties due to incompatibility between the elastic global analysis and the limit state member design in the conventional LRFD method. The genetic algorithm uses a procedure based on Darwinian notions of survival of the fittest, where selection, crossover, and mutation operators are used among sections in the database to look for high performance ones. They satisfy the constraint functions and give the lightest weight to the structure. The objective function is set to the total weight of the steel structure and the constraint functions are load-carrying capacities, serviceability, and ductility requirement. Case studies of a three-dimensional frame and a three-dimensional steel arch bridge are presented.

• Journal of the Earthquake Engineering Society of Korea
• /
• v.19 no.5
• /
• pp.239-246
• /
• 2015

Reinforced concrete shear walls are effective for resisting lateral loads imposed by wind or earthquakes. Observed damages of the shear wall in recent earthquakes in Chile(2010) and New Zealand(2011) exceeded expectations. Various analytical models have been proposed in order to incorporate such response features in predicting the inelastic response of RC shear walls. However, the model has not been implemented into widely available computer programs, and has not been sufficiently calibrated with and validated against extensive experimental data at both local and global response levels. In this study, reinforced concrete shear walls were modeled with fiber slices, where cross section and reinforcement details of shear walls can be arranged freely. Nonlinear analysis was performed by adding nonlinear shear spring elements that can represent shear deformation. This analysis result will be compared with the existing experiment results. To investigate the nonlinear behavior of reinforced concrete shear walls, reinforced concrete single shear walls with rectangular wall cross section were selected. The analysis results showed that the yield strength of the shear wall was approximately the same value as the experimental results. However, the yielding displacement of the shear wall was still higher in the experiment than the analysis. The analytical model used in this study is available for the analysis of shear wall subjected to high axial forces.

• Proceedings of the Computational Structural Engineering Institute Conference
• /
• 2011.04a
• /
• pp.278-281
• /
• 2011

The global coordinates generated from Isomap algorithm provide a simple way to analyze and manipulate high dimensional observations in terms of their intrinsic nonlinear degrees of freedom. Thus, Isomap can find globally meaningful coordinates and nonlinear structure of complex data sets, while neither principal component analysis (PCA) nor multidimensional scaling (MDS) are successful in many cases. It is demonstrated that the adapted Isomap algorithm successfully enhances the quality of pattern classification for damage identification in various numerical examples.

• Proceedings of the Korean Society for Noise and Vibration Engineering Conference
• /
• 2004.11a
• /
• pp.140-144
• /
• 2004

Global bifurcations and chaos in modal interactions of an imperfect circular plate with one-to-one internal resonance are investigated. The case of primary resonance, in which an excitation frequency is near natural frequencies, is considered. The damping force is not included in the analysis. The Melnikov's method for heteroclinic orbits of the autonomous system was used to obtain the criteria for chaotic motion.