• The Transactions of The Korean Institute of Electrical Engineers
• /
• v.56 no.7
• /
• pp.1315-1320
• /
• 2007

In this paper, we introduce a new fuzzy model called fuzzy combined polynomial neural networks, which are based on the representative fuzzy model named polynomial fuzzy model. In the design procedure of the proposed fuzzy model, the coefficients on consequent parts are estimated by using not general least square estimation algorithm that is a sort of global learning algorithm but weighted least square estimation algorithm, a sort of local learning algorithm. We are able to adopt various type of structures as the consequent part of fuzzy model when using a local learning algorithm. Among various structures, we select Polynomial Neural Networks which have nonlinear characteristic and the final result of which is a complex mathematical polynomial. The approximation ability of the proposed model can be improved using Polynomial Neural Networks as the consequent part.

• Steel and Composite Structures
• /
• v.22 no.2
• /
• pp.277-299
• /
• 2016

In this paper, free vibration, forced vibration, resonance and stress wave propagation behavior in nanocomposite plates reinforced by wavy carbon nanotube (CNT) are studied by a mesh-free method based on first order shear deformation theory (FSDT). The plates are resting on Winkler-Pasternak elastic foundation and subjected to periodic or impact loading. The distributions of CNTs are considered functionally graded (FG) or uniform along the thickness and their mechanical properties are estimated by an extended rule of mixture. In the mesh-free analysis, moving least squares (MLS) shape functions are used for approximation of displacement field in the weak form of motion equation and the transformation method is used for imposition of essential boundary conditions. Effects of CNT distribution, volume fraction, aspect ratio and waviness, and also effects of elastic foundation coefficients, plate thickness and time depended loading are examined on the vibrational and stresses wave propagation responses of the nanocomposite plates reinforced by wavy CNT.

• Proceedings of the KOSOMBE Conference
• /
• v.1997 no.05
• /
• pp.184-187
• /
• 1997

In terms of a system identification, we present a method for autoregressive(AR) modelling of variious biomedical signal. Model order is estimated fly low rank approximation and coefficients are determined by innovation processes of Kalman filter derivation. An application of the method is given for visual evoked potentials.

• Proceedings of the KIEE Conference
• /
• 1999.11c
• /
• pp.797-799
• /
• 1999

In this paper, we design the Multi-FNN(Fuzzy-Neural Networks) using HCM Method. The proposed Multi-FNN uses simplified inference as fuzzy inference method and Error Back Propagation Algorithm as learning rules. Also, We use HCM(Hard C-Means) method of clustering technique for improvement of output performance from pre-processing of input data. The parameters such as apexes of membership function, learning rates and momentum coefficients are adjusted using genetic algorithms. We use the training and testing data set to obtain a balance between the approximation and the generalization of our model. Several numerical examples are used to evaluate the performance of the our model. From the results, we can obtain higher accuracy and feasibility than any other works presented previously.

• Proceedings of the KIEE Conference
• /
• 1999.11c
• /
• pp.824-826
• /
• 1999

In this paper, we design an Fuzzy-Neural Networks(FNN) by means of divisions of fuzzy input space with multi-input variables. Fuzzy input space of Yamakawa's FNN is divided by each separated input variable, but that of the proposed FNN is divided by mutually combined input variables. The membership functions of the proposed FNN use both triangular and gaussian membership types. The parameters such as apexes of membership functions, learning rates, momentum coefficients, weighting value, and slope are adjusted using genetic algorithms. Also, an aggregate objective function(performance index) with weighting value is utilized to achieve a sound balance between approximation and generalization abilities of the model. To evaluate the performance of the proposed model, we use the data of sewage treatment process.

• The Journal of Korean Institute of Communications and Information Sciences
• /
• v.24 no.6A
• /
• pp.881-887
• /
• 1999

In this paper, were introduce two types of equalizers, called equalizer-a and equalizer-b, applying to wireless communications having unknown channel characteristics. The equalizer-a, which has the single sample detector with equalizer system, is developed while the equalizer-b has the partition detectors with the same system used in equalizer-a. The methodologiy we adopt for designing the equalizers is that the sample space is partitioned into finite number of regions by using quantiles, which are estimated by robbins-monro stochastic approximation (RMSA) algorithm, and the coefficients of equalizers are calculated based on nonlinear minimum mean, square error (MMSE) algorithm. Through the computer simulation, the equalizers show much better performance in equiprobably partitioned sample subspaces of observations than the single sample detector and the detector, which has the conventional equalizer, in unquantized observation space under various noise environments.

• Proceedings of the KIEE Conference
• /
• 2001.11c
• /
• pp.420-423
• /
• 2001

In this paper, the Multi-FNNs(Fuzzy-Neural Networks) architecture is identified and optimized using HCM(Hard C-Means) clustering method and genetic algorithms. The proposed Multi-FNNs architecture uses simplified inference and linear inference as fuzzy inference method and error back propagation algorithm as learning rules. Here, HCM clustering method, which is carried out for the process data preprocessing of system modeling, is utilized to determine the structure of Multi-FNNs according to the divisions of input-output space using I/O process data. Also, the parameters of Multi-FNNs model such as apexes of membership function, learning rates and momentum coefficients are adjusted using genetic algorithms. An aggregate performance index with a weighting factor is used to achieve a sound balance between approximation and generalization abilities of the model. To evaluate the performance of the proposed model we use the time series data for gas furnace and the NOx emission process data of gas turbine power plant.

• Proceedings of the KIEE Conference
• /
• 2000.07e
• /
• pp.17-21
• /
• 2000

A Study on the electron energy distribution function in $SF_6+Ar$ mixtures gas used by MCS-BE algorithm, the electron swam parameters in the 0.5% and 0.2% $SF_6+Ar$ mixtures are measured by time of flight method over the E/N(Td) range from 30 to 300(Td). A two-term approximation of the Boltzmann equation analysis and Monte Carlo simulation have been also used to study electron transport coefficients. The electron energy distribution function has been analysed in $SF_6$ gas and $SF_6+Ar$ mixtures at E/N : 200(Td) for a case of the equilibrium region in the mean electron energy. The measured results and the calculated results have been compared each other.

• Proceedings of the KIEE Conference
• /
• 2004.07e
• /
• pp.37-40
• /
• 2004

Electron swarm parameters in pure $CF_4$ and mixtures of $CF_4$ and Ar, have been analyzed over a range of the reduced electric field strength between 0.1 and 350[Td] by the two-term approximation of the Boltzmann equation(BEq.) method and the Monte Carlo simulation(MCS) The results of the Boltzmann equation and the Monte Carlo simulation have been compared with the data presented by several workers. The deduced transport coefficients for electrons agree reasonably well with the experimental and simulation data obtained by Nakamura and Hayashi. The energy distribution function of electrons in $CF_4$-Ar mixtures shows the Maxwellian distribution for energy. That is, f(${\varepsilon}$) has the symmetrical shape whose axis of symmetry is a most probably energy

• 제어로봇시스템학회:학술대회논문집
• /
• 2002.10a
• /
• pp.97.2-97
• /
• 2002

A method for the non-destructive detection of damage using parameterized partial differential equations and Galerkin approximation techniques is presented. This method provides the theoretical and experimental verification of a nondestructive time domain approach to examine structural damage in smart structure. The time histories of the vibration response of structure were used to identify the presence of damage. Damage in a structure causes changes in the physical coefficients of mass density, elastic modulus and damping coefficient. This paper examines the beam-like structures with PVDF sensor and PZT actuator to perform identification of those physical parameters and to detect the...