• Journal of Korean Society of Steel Construction
• /
• v.16 no.1 s.68
• /
• pp.123-134
• /
• 2004

In this research, a multilevel decomposition technique to enhance the efficiency of the configuration optimization of truss structures was proposed. On the first level, the nonlinear programming problem was formulated considering cross-sectional areas as design variables, weight, or volume as objective function and behavior under multiloading condition as design constraint. Said nonlinear programming problem was transformed into a sequential linear programming problem. which was effective in calculation through the approximation of member forces using behavior space approach. Such approach has proven to be efficient in sensitivity analysis and different form existing shape optimization studies. The modified method of feasible direction (MMFD) was used for the optimization process. On the second level, by treating only shape design variables, the optimum problem was transformed into and unconstrained optimal design problem. A unidirectional search technique was used. As numerical examples, some truss structures were applied to illustrate the applicability. and validity of the formulated algorithm.

• Proceedings of the Korean Institute of Intelligent Systems Conference
• /
• 2004.10a
• /
• pp.433-436
• /
• 2004

In this paper, we introduce a new Fuzzy Polynomial Neural Networks (FPNNS)-like structure whose neuron is based on the Fuzzy Set-based Fuzzy Inference System (FS-FIS) and is different from that of FPNNS based on the Fuzzy relation-based Fuzzy Inference System (FR-FIS) and discuss the ability of the new FPNNS-like structure named Fuzzy Set-based Polynomial Neural Networks (FSPNN). The premise parts of their fuzzy rules are not identical, while the consequent parts of the both Networks (such as FPNN and FSPNN) are identical. This difference results from the angle of a viewpoint of partition of input space of system. In other word, from a point of view of FS-FIS, the input variables are mutually independent under input space of system, while from a viewpoint of FR-FIS they are related each other. The proposed design procedure for networks architecture involves the selection of appropriate nodes with specific local characteristics such as the number of input variables, the order of the polynomial that is constant, linear, quadratic, or modified quadratic functions being viewed as the consequent part of fuzzy rules, and a collection of the specific subset of input variables. On the parameter optimization phase, we adopt Information Granulation (IC) based on HCM clustering algorithm and a standard least square method-based learning. Through the consecutive process of such structural and parametric optimization, an optimized and flexible fuzzy neural network is generated in a dynamic fashion. To evaluate the performance of the genetically optimized FSPNN (gFSPNN), the model is experimented with using the time series dataset of gas furnace process.

• Journal of the Korean Society for Aeronautical & Space Sciences
• /
• v.36 no.11
• /
• pp.1039-1048
• /
• 2008

new computational procedure for the Non-Linear Vortex Lattice Method (NLVLM) is suggested in this work. Conventional procedures suggested so far usually involves inner iteration loop to update free vortex shape and an under-relaxation based iteration loop to determine the free vortex shape. In this present work, we suggest a new formula based on quasi-steady concept to fix free vortex shape which eliminates the need for inner iteration loop. Further, the ensemble averaging of the induced velocities for a given free vortex segment evaluated at each iteration significantly improves the convergence property of the algorithm without resorting to the under-relaxation technique. Numerical experiments over several low aspect ratio wings are carried out to obtain optimal empirical parameters such as the length of the free vortex segment, the vortex core radius, and the rolled-up wake length.

• The Journal of Korean Institute of Communications and Information Sciences
• /
• v.27 no.7A
• /
• pp.641-650
• /
• 2002

Space-Time Block Code(STBC) provides full diversity gains with simple linear processing at the receiver. Interference Cancellation(IC) techniques in system using STBC improve the capacity and performance of wireless systems with co-channel users. Various IC techniques, Minimum Mean-Squared Error(MMSE) and Zero-Forcing(ZF) algorithms in system with STBC were proposed in the literatures in multiuser environment. The performance of these IC techniques were simulated by assuming perfect channel state information(CSI) of multiuser at the receiver. However, in practice it is difficult to know perfect CSI of multiuser at the receiver. Thus, channel estimation scheme is essential at the receiver. Also SNR estimation scheme is required to operate the MMSE IC algorithm. In this paper, we present estimation schemes of CSI and SNR using training sequences. Through extensive computer simulation, we compare and evaluate the performance of IC techniques using the proposed CSI and SNR estimation techniques.

• The Transactions of The Korean Institute of Electrical Engineers
• /
• v.66 no.2
• /
• pp.416-424
• /
• 2017

In this study, we propose a design of digits recognition system based on RBFNNs through a comparative study of pre-processing algorithms in order to recognize digits in handwritten. Histogram of Oriented Gradient(HOG) is used to get the features of digits in the proposed digits recognition system. In the pre-processing part, a dimensional reduction is executed by using Principal Component Analysis(PCA) and (2D)2PCA which are widely adopted methods in order to minimize a loss of the information during the reduction process of feature space. Also, The architecture of radial basis function neural networks consists of three functional modules such as condition, conclusion, and inference part. In the condition part, the input space is partitioned with the use of fuzzy clustering realized by means of the Fuzzy C-Means algorithm. Also, it is used instead of gaussian function to consider the characteristic of input data. In the conclusion part, the connection weights are used as the extended type of polynomial expression such as constant, linear, quadratic and modified quadratic. By using MNIST handwritten digit benchmarking database, experimental results show the effectiveness and efficiency of proposed digit recognition system when compared with other studies.

• Journal of the Korea Institute of Information and Communication Engineering
• /
• v.11 no.8
• /
• pp.1596-1603
• /
• 2007

In this paper, we experimented by three kind of method for feature parameter, training method and recognition algorithm of most suitable for speech recognition system and considered. We decided speech recognition system of most suitable through two kind of experiment after we make speech recognizer. First, we did an experiment about three kind of feature parameter to evaluate recognition performance of it in speech recognizer using existent MFCC and MFCC new feature parameter that change characteristic space using PCA and ICA. Second, we experimented recognition performance or HMM, SVM and RVM by studying data number. By an experiment until now, feature parameter by ICA showed performance improvement of average 1.5% than MFCC by high linear discrimination from characteristic space. RVM showed performance improvement of maximum 3.25% than HMM in an experiment by decrease of studying data. As such result, effective method for speech recognition system to propose in this paper derives feature parameters using ICA and un recognition using RVM.

• KSCE Journal of Civil and Environmental Engineering Research
• /
• v.26 no.3B
• /
• pp.279-289
• /
• 2006

The reconstruction of low dimension nonlinear behavior from the hydrologic time series has been an active area of research in the last decade. In this study, we present the applications of a powerful state space reconstruction methodology using the method of Support Vector Machines (SVM) to the Great Salt Lake (GSL) volume. SVMs are machine learning systems that use a hypothesis space of linear functions in a Kernel induced higher dimensional feature space. SVMs are optimized by minimizing a bound on a generalized error (risk) measure, rather than just the mean square error over a training set. The utility of this SVM regression approach is demonstrated through applications to the short term forecasts of the biweekly GSL volume. The SVM based reconstruction is used to develop time series forecasts for multiple lead times ranging from the period of two weeks to several months. The reliability of the algorithm in learning and forecasting the dynamics is tested using split sample sensitivity analyses, with a particular interest in forecasting extreme states. Unlike previously reported methodologies, SVMs are able to extract the dynamics using only a few past observed data points (Support Vectors, SV) out of the training examples. Considering statistical measures, the prediction model based on SVM demonstrated encouraging and promising results in a short-term prediction. Thus, the SVM method presented in this study suggests a competitive methodology for the forecast of hydrologic time series.

• Journal of Navigation and Port Research
• /
• v.35 no.1
• /
• pp.83-91
• /
• 2011

The accuracy of forecasting is remarkably important to reduce total cost or to increase customer services, so it has been studied by many researchers. In this paper, the artificial neural network (ANN), one of the most popular nonlinear forecasting methods, is compared with autoregressive integrated moving average(ARIMA) model through performing a prediction of container traffic. It uses a hybrid methodology that combines both the linear ARIAM and the nonlinear ANN model to improve forecasting performance. Also, it compares the methodology with other models in performance for prediction. In designing network structure, this work specially applies the genetic algorithm which is known as the effectively optimal algorithm in the huge and complex sample space. It includes the time delayed neural network (TDNN) as well as multi-layer perceptron (MLP) which is the most popular neural network model. Experimental results indicate that both ANN and Hybrid models outperform ARIMA model.

• Tunnel and Underground Space
• /
• v.13 no.6
• /
• pp.434-443
• /
• 2003

In this study, it was purposed to develop the new method for the prediction of pollutant concentration in road tunnels. The new method was the use of artificial neural network with the back-propagation algorithm which can model the non-linear system of tunnel environment. This network system was separated into two parts as the visibility and the CO concentration. For this study, data was collected from two highway road tunnels on Yeongdong Expressway. The tunnels have two lanes with one-way direction and adopt the longitudinal ventilation system. The actually measured data from the tunnels was used to develop the neural network system for the prediction of pollutant concentration. The output results from the newly developed neural network system were analysed and compared with the calculated values by PIARC method. Results showed that the prediction accuracy by the neural network system was approximately five times better than the one by PIARC method. In addition, the system predicted much more accurately at the situation where the drivers have to be stayed for a while in tunnels caused by the low velocity of vehicles.

• Journal of Korean Society of Transportation
• /
• v.21 no.6
• /
• pp.77-88
• /
• 2003

This paper proposes a solution algorithm for solving a multi-class analytical DTA model. In the DTA model, three traveler classes are classified according to different assumptions of traveler's route choice behavior: including fixed route, Stochastic Dynamic User Optimum(SDUO), and Dynamic User Optimum(DUO). To solve this DTA model, variables of link flow and exit flow are represented solely by inflow. The resulting Linear Program(LP) subproblem in the inner iteration is solved as a typical time-dependent shortest route problem over a physical network. Accordingly, the required time-space network expansion in solving DTA models is no longer needed.