• 제어로봇시스템학회:학술대회논문집
• /
• 1993.10a
• /
• pp.14-19
• /
• 1993

In this paper, we proposed an indirect learning and direct adaptive control schemes using neural networks, i.e., composite adaptive neural control, for a class of continuous nonlinear systems. With the indirect learning method, the neural network learns the nonlinear basis of the system inverse dynamics by a modified backpropagation learning rule. The basis spans the local vector space of inverse dynamics with the direct adaptation method when the indirect learning result is within a prescribed error tolerance, as such this method is closely related to the adaptive control methods. Also hash addressing technique, similar to the CMAC functional architecture, is introduced for partitioning network hidden nodes according to the system states, so global neuro control properties can be organized by the local ones. For uniform stability, the sliding mode control is introduced when the neural network has not sufficiently learned the system dynamics. With proper assumptions on the controlled system, global stability and tracking error convergence proof can be given. The performance of the proposed control scheme is demonstrated with the simulation results of a nonlinear system.

• The Transactions of the Korean Institute of Electrical Engineers
• /
• v.39 no.9
• /
• pp.985-996
• /
• 1990

An inverse kinematic solution of a robot manipulator using multilayer perceptrons is proposed. Neural networks allow the solution of some complex nonlinear equations such as the inverse kinematics of a robot manipulator without the need for its model. However, the back-propagation (BP) learning rule for multilayer perceptrons has the major limitation of being too slow in learning to be practical. In this paper, a new algorithm named Dynamically Reconfiguring BP is proposed to improve its learning speed. It uses a modified version of Kohonen's Self-Organizing Feature Map (SOFM) to partition the input space and for each input point, select a subset of the hidden processing elements or neurons. A subset of the original network results from these selected neuron which learns the desired mapping for this small input region. It is this selective property that accelerates convergence as well as enhances resolution. This network was used to learn the parity function and further, to solve the inverse kinematic problem of a robot manipulator. The results demonstrate faster learning than the BP network.

• Journal of the Korean Society for Aeronautical & Space Sciences
• /
• v.30 no.7
• /
• pp.44-50
• /
• 2002

Low-velocity impact on composite sandwich panel has been investigated. Contact force is computed from a proposed modified Hertzian contact law. The Hertzian contact law is constructed by adjusting numerical value of the exponent and reducing the through-the- thickness elastic constant of honeycomb core. The equivalent transverse elastic constant is calculated from the rule of mixture. Nonlinear equation to calculate the contact force is solved by the Newton-Raphson method and time integration is done by the Newmark-beta method. A finite element program for the low-velocity impact analysis is coded by implementing these techniques and an 18-node assumed strain solid element. Behaviors of composite sandwich panels subjected to low-velocity impact are analyzed for various cases with different geometry and lay-ups. It has been found that the present code with the proposed contact law can predict measured contact forces and contact times for most cases within reasonable error bounds.

• Journal of Electrical Engineering and Technology
• /
• v.7 no.4
• /
• pp.636-645
• /
• 2012

This study introduces an information granular-based fuzzy radial basis function neural networks (FRBFNN) based on multiobjective optimization and weighted least square (WLS). An improved multiobjective space search algorithm (IMSSA) is proposed to optimize the FRBFNN. In the design of FRBFNN, the premise part of the rules is constructed with the aid of Fuzzy C-Means (FCM) clustering while the consequent part of the fuzzy rules is developed by using four types of polynomials, namely constant, linear, quadratic, and modified quadratic. Information granulation realized with C-Means clustering helps determine the initial values of the apex parameters of the membership function of the fuzzy neural network. To enhance the flexibility of neural network, we use the WLS learning to estimate the coefficients of the polynomials. In comparison with ordinary least square commonly used in the design of fuzzy radial basis function neural networks, WLS could come with a different type of the local model in each rule when dealing with the FRBFNN. Since the performance of the FRBFNN model is directly affected by some parameters such as e.g., the fuzzification coefficient used in the FCM, the number of rules and the orders of the polynomials present in the consequent parts of the rules, we carry out both structural as well as parametric optimization of the network. The proposed IMSSA that aims at the simultaneous minimization of complexity and the maximization of accuracy is exploited here to optimize the parameters of the model. Experimental results illustrate that the proposed neural network leads to better performance in comparison with some existing neurofuzzy models encountered in the literature.

• Journal of the Korean Institute of Intelligent Systems
• /
• v.18 no.5
• /
• pp.685-691
• /
• 2008

The selection of classification boundaries in fuzzy rule- based classification systems is an important and difficult problem. So various methods based on learning processes such as neural network, genetic algorithm, and so on have been proposed for it. In a previous study, we pointed out the limitation of the methods and discussed a method for fuzzy partitioning in the overlapped region on feature space in order to overcome the time-consuming when the additional parameters for tuning fuzzy membership functions are necessary. In this paper, we propose a method to determine three types of classification boundaries(i.e., non-overlapping, overlapping, and a boundary point) on the basis of statistical information of the given dataset without learning by extending the method described in the study. Finally, we show the effectiveness of the proposed method through experimental results applied to pattern classification problems using the modified IRIS and standard IRIS datasets.

• Transactions of the Korean Society of Mechanical Engineers B
• /
• v.21 no.11
• /
• pp.1538-1551
• /
• 1997

Kinetic energy conservation for fixed and moving grids is examined in time-accurate finite element computation of fully unsteady inviscid flows. As numerical algorithms, fractional step method (FSM) and modified SIMPLE are used. To simulate the flow in moving grid system, arbitrary Lagrangian-Eulerian (ALE) method is adopted. In the present study, the energy conserving time integration rule for finite element algorithm is proposed and discussed schematically. It is shown that the discretization by Crank-Nicolson in time and Galerkin (central difference) in space must be used to ensure energy conservation. The developed code has been tested for a standing vortex in fixed or moving grid system, sloshing in a tank and propagation of a solitary wave, and has been shown to be a completely energy conserving algorithm.

• Structural Engineering and Mechanics
• /
• v.85 no.2
• /
• pp.147-161
• /
• 2023

Based on the ideas of variational differential quadrature (VDQ) and finite element method (FEM), a numerical approach named as VDQFEM is applied herein to study the large deformations of plate-type structures under static loading with arbitrary shape hole made of functionally graded graphene platelet-reinforced composite (FG-GPLRC) in the context of higher-order shear deformation theory (HSDT). The material properties of composite are approximated based upon the modified Halpin-Tsai model and rule of mixture. Furthermore, various FG distribution patterns are considered along the thickness direction of plate for GPLs. Using novel vector/matrix relations, the governing equations are derived through a variational approach. The matricized formulation can be efficiently employed in the coding process of numerical methods. In VDQFEM, the space domain of structure is first transformed into a number of finite elements. Then, the VDQ discretization technique is implemented within each element. As the last step, the assemblage procedure is performed to derive the set of governing equations which is solved via the pseudo arc-length continuation algorithm. Also, since HSDT is used herein, the mixed formulation approach is proposed to accommodate the continuity of first-order derivatives on the common boundaries of elements. Rectangular and circular plates under various boundary conditions with circular/rectangular/elliptical cutout are selected to generate the numerical results. In the numerical examples, the effects of geometrical properties and reinforcement with GPL on the nonlinear maximum deflection-transverse load amplitude curve are studied.

• Geomechanics and Engineering
• /
• v.29 no.6
• /
• pp.615-626
• /
• 2022

The soft clays are widely distributed, and one of the prominent engineering problems is the creep behavior. In order to predict the creep deformation of soft clays in an easier and more acceptable way, a simple creep constitutive model has been proposed in this paper. Firstly, the triaxial creep test data indicated that, the strain-time (𝜀-t) curve showing in the 𝜀-lgt space can be divided into two lines with different slopes, and the time referring to the demarcation point is named as t_EOP. Thereafter, the strain increments occurred after the time t_EOP are totally assumed to be the creep components, and the elastic and plastic strains had occurred before t_EOP. A hyperbolic equation expressing the relationship between creep volumetric strain, stress and time is proposed, with several triaxial creep test data of soft clays verifying the applicability. Additionally, the creep flow law is suggested to be similar with the plastic flow law of the modified Cam-Clay model, and the proposed volumetric strain equation is used to deduced the scaling factor for creep strains. Therefore, a creep constitutive model is thereby established, and verified by successfully predicting the creep principal strains of triaxial specimens.

• Journal of Astronomy and Space Sciences
• /
• v.27 no.1
• /
• pp.1-10
• /
• 2010

The distance distribution in our planetary system has been a controversial matter. Two kinds of important issues on Titius-Bode's relation have been discussed up to now: one is if there is a simple mathematical relation between distances of natural bodies orbiting a central body, and the other is if there is any physical basis for such a relation. We have examined, by applying it to exo-planetary systems, whether Titius-Bode's relation is exclusively applicable to our solar system. We study, with the $X^2$ test, the distribution of period ratios of two planets in multiple planet systems by comparing it with that derived from not only Titius-Bode's relation but also other forms of it. The $X^2$ value between the distribution of the orbital period derived from Titius-Bode's relation and that observed in our Solar system is 12.28 (dof=18) with high probability, i.e., 83.3 %. The value of $X^2$ and probability resulted from Titius-Bode's relation and observed exo-planetary systems are 21.38 (dof=26) and 72.2 %, respectively. Modified forms we adopted seem also to agree with the planetary system as favorably as Titius-Bode's relation does. As a result, one cannot rule out the possibility that the distribution of the ratio of orbiting periods in multiple planet systems is consistent with that derived from Titius-Bode's relation. Having speculated Titius-Bode's relation could be valid in exo-planetary systems, we tentatively conclude it is unlikely that Titius-Bode's relation explains the distance distribution in our planetary system due to chance. Finally, we point out implications of our finding.

• Journal of Korean Tunnelling and Underground Space Association
• /
• v.17 no.2
• /
• pp.65-74
• /
• 2015

In order to design an advanced composite materials drainage pipe structures for an undersea tunnel, mechanical properties for the lamina types of the structural member must be predetermined. It is also reported that the size effect of the specimen is significant. In this study the tensile tests for the lamina types of the structural member are conducted at the room temperature ($20^{\circ}C$) and the seawater temperature ($0^{\circ}C$). In addition, the mechanical properties are predicted by theory based on the rule of mixtures and elasticity solution technique. The predicted mechanical properties are compared with test results obtained by a test method. In the design of an advanced composite materials drainage pipe structural members for an undersea tunnel, the used mechanical properties must be applied at the room temperature with considering the modified factors. These are to be offered the datum for the design an advanced composite materials drainage pipe structures for an undersea tunnel.