• Journal of the Institute of Electronics Engineers of Korea SP
• /
• v.42 no.5 s.305
• /
• pp.47-54
• /
• 2005

In general, wavelet coefficients are composed of a few large coefficients and a lot of small coefficients. In this paper, we propose image denoising algorithm using Bernoulli-Gaussian mixture model based on sparse characteristic of wavelet coefficient. The Bernoulli-Gaussian mixture is composed of the multiplication of Bernoulli random variable and Gaussian mixture random variable. The image denoising is performed by using Bayesian estimation. We present an effective denoising method through simplified parameter estimation for Bernoulli random variable using local expected squared error. Simulation results show our method outperforms the states-of-art denoising methods when using orthogonal wavelets.

• Journal of Power Electronics
• /
• v.14 no.5
• /
• pp.1008-1027
• /
• 2014

An electric scooter with a V-belt continuously variable transmission (CVT) driven by a permanent magnet synchronous motor (PMSM) has a lot of nonlinear and time-varying characteristics, and accurate dynamic models are difficult to establish for linear controller designs. A PMSM servo-drive electric scooter controlled by a novel hybrid modified recurrent Legendre neural network (NN) control system is proposed to solve difficulties of linear controllers under the occurrence of nonlinear load disturbances and parameters variations. Firstly, the system structure of a V-belt CVT driven electric scooter using a PMSM servo drive is established. Secondly, the novel hybrid modified recurrent Legendre NN control system, which consists of an inspector control, a modified recurrent Legendre NN control with an adaptation law, and a recouped control with an estimation law, is proposed to improve its performance. Moreover, the on-line parameter tuning method of the modified recurrent Legendre NN is derived according to the Lyapunov stability theorem and the gradient descent method. Furthermore, two optimal learning rates for the modified recurrent Legendre NN are derived to speed up the parameter convergence. Finally, comparative studies are carried out to show the effectiveness of the proposed control scheme through experimental results.

• Structural Engineering and Mechanics
• /
• v.28 no.6
• /
• pp.749-765
• /
• 2008

Free axisymmetric vibrations of layered cylindrical shells of variable thickness are studied using spline function approximation techniques. Three different types of thickness variations are considered namely linear, exponential and sinusoidal. The equations of axisymmetric motion of layered cylindrical shells, on the longitudinal and transverse displacement components are obtained using Love's first approximation theory. A system of coupled differential equations on displacement functions are obtained by assuming the displacements in a separable form. Then the displacements are approximated using Bickley-spline approximation. The vibrations of two-layered cylindrical shells, made up of several types of layered materials and different boundary conditions are considered. Parametric studies have been made on the variation of frequency parameter with respect to the relative layer thickness, length ratio and type of thickness variation parameter.

• Advances in nano research
• /
• v.17 no.3
• /
• pp.235-248
• /
• 2024

This study explores the transmission of waves through polymer composite nanoplates situated on varying elastic foundations. The reinforcement of these nanoplates is assured by graphene nanoplatelets (GNP). Furthermore, the material's behavior is assessed using the Halpin-Tsai model, while the precise representations of stress and strain effects are ensured by the four variables higher order shear deformation theory. The equations of motion are obtained and resolved through the application of Hamilton's principle and the trial function. The study examines how different factors, like the nonlocal parameter, strain gradient parameter, weight fraction, and variable elastic foundations affect the outcomes of wave propagation in nanoplates. This thorough investigation offers valuable insights into the difficult behavior of wave dynamics in nanoplates, this has led to substantial advancements in engineering applications for the future.

• Steel and Composite Structures
• /
• v.50 no.5
• /
• pp.563-581
• /
• 2024

This paper presents a novel finite element model for the free vibration analysis of variable-thickness functionally graded porous (FGP) microplates resting on Pasternak's medium in the hygro-thermal environment. The governing equations are established according to refined higher-order shear deformation plate theory (RPT) in construction with the modified couple stress theory. For the first time, three-node triangular elements with twelve degrees of freedom for each node are developed based on Hermitian interpolation functions to describe the in-plane displacements and transverse displacements of microplates. Two laws of variable thickness of FGP microplates, including the linear law and the nonlinear law in the x-direction are investigated. Effects of thermal and moisture changes on microplates are assumed to vary continuously from the bottom surface to the top surface and only cause tension loads in the plane, which does not change the material's mechanical properties. The numerical results of this work are compared with those of published data to verify the accuracy and reliability of the proposed method. In addition, the parameter study is conducted to explore the effects of geometrical and material properties such as the changing law of the thickness, length-scale parameter, and the parameters of the porosity, temperature, and humidity on the free vibration response of variable thickness FGP microplates. These results can be applied to design of microelectromechanical structures in practice.

• KSII Transactions on Internet and Information Systems (TIIS)
• /
• v.7 no.1
• /
• pp.81-98
• /
• 2013

Since Probabilistic Latent Semantic Analysis (PLSA) and Latent Dirichlet Allocation (LDA) were introduced, many revised or extended topic models have appeared. Due to the intractable likelihood of these models, training any topic model requires to use some approximation algorithm such as variational approximation, Laplace approximation, or Markov chain Monte Carlo (MCMC). Although these approximation algorithms perform well, training a topic model is still computationally expensive given the large amount of data it requires. In this paper, we propose a new method, called non-simultaneous sampling deactivation, for efficient approximation of parameters in a topic model. While each random variable is normally sampled or obtained by a single predefined burn-in period in the traditional approximation algorithms, our new method is based on the observation that the random variable nodes in one topic model have all different periods of convergence. During the iterative approximation process, the proposed method allows each random variable node to be terminated or deactivated when it is converged. Therefore, compared to the traditional approximation ways in which usually every node is deactivated concurrently, the proposed method achieves the inference efficiency in terms of time and memory. We do not propose a new approximation algorithm, but a new process applicable to the existing approximation algorithms. Through experiments, we show the time and memory efficiency of the method, and discuss about the tradeoff between the efficiency of the approximation process and the parameter consistency.

• Steel and Composite Structures
• /
• v.34 no.4
• /
• pp.511-524
• /
• 2020

In this research, a simple four-variable trigonometric integral shear deformation model is proposed for the static behavior of advanced functionally graded (AFG) ceramic-metal plates supported by a two-parameter elastic foundation and subjected to a nonlinear hygro-thermo-mechanical load. The elastic properties, including both the thermal expansion and moisture coefficients of the plate, are also supposed to be varied within thickness direction by following a power law distribution in terms of volume fractions of the components of the material. The interest of the current theory is seen in its kinematics that use only four independent unknowns, while first-order plate theory and other higher-order plate theories require at least five unknowns. The "in-plane displacement field" of the proposed theory utilizes cosine functions in terms of thickness coordinates to calculate out-of-plane shear deformations. The vertical displacement includes flexural and shear components. The elastic foundation is introduced in mathematical modeling as a two-parameter Winkler-Pasternak foundation. The virtual displacement principle is applied to obtain the basic equations and a Navier solution technique is used to determine an analytical solution. The numerical results predicted by the proposed formulation are compared with results already published in the literature to demonstrate the accuracy and efficiency of the proposed theory. The influences of "moisture concentration", temperature, stiffness of foundation, shear deformation, geometric ratios and volume fraction variation on the mechanical behavior of AFG plates are examined and discussed in detail.

• Nuclear Engineering and Technology
• /
• v.17 no.4
• /
• pp.247-255
• /
• 1985

This paper describes a new method for the design of variable structure model-following control systems(VSMFC). This design concept is developed using the theory of variable structure systems (VSS) and slide mode. The new results are presented on the sliding control methodology to achieve accurate tracking for a class of nonlinear, multi-input multi-output(MIMO), time varying systems in the presence of parameter variations. The design requires little computational effort. The dynamic response is insensitive to parameter variations. The feasibility and the advantages of the method are illustrated by applying it to a 1000 MWe boiling water reactor(BWR). The control is studied in the range of 85%∼90% of rated power for load-following control. A set of 12 nonlinear differential equations is used to simulate the total plant. A 6-th order linear model has been developed from these equations at 85% of rated power. The obtained controller is shown by simulations to be able to compensate for a plant parameter variation over a wide power range.

• Proceedings of the KIEE Conference
• /
• 1990.11a
• /
• pp.227-230
• /
• 1990

In this paper we present a variable structure systeme-model following adaptive control (VSS-MFAC) method for an uncertain turbo-generator system which is apt to suffer from the unmodeled parameter uncertainties and the external disturbances. The simulation results for the power system stabilizer(PSS) exhibit robust adaptive model-following properties well in the PSS designed by the proposed VSS-MFAC methodology when a step change in the mechanical torque and a parameter variation is applied.

• Journal of the Korean Data and Information Science Society
• /
• v.12 no.2
• /
• pp.57-64
• /
• 2001

In the Taguchi parameter design, the product-array approach using orthogonal arrays is mainly used. However, it often requires an excessive number of experiments. An alternative approach, which is called the combined-array approach, was suggested by Welch et al (1990) and studied by Vining and Myers (1990) and others. In these studies, only single respouse variable was considered. We propose how to simultaneously optimize multiple responses when there are correlations among responses.