• The Transactions of The Korean Institute of Electrical Engineers
• /
• v.60 no.3
• /
• pp.639-647
• /
• 2011

In this paper, we introduce a design methodology of data-centroid Radial Basis Function neural networks with extended polynomial function. The two underlying design mechanisms of such networks involve K-means clustering method and Particle Swarm Optimization(PSO). The proposed algorithm is based on K-means clustering method for efficient processing of data and the optimization of model was carried out using PSO. In this paper, as the connection weight of RBF neural networks, we are able to use four types of polynomials such as simplified, linear, quadratic, and modified quadratic. Using K-means clustering, the center values of Gaussian function as activation function are selected. And the PSO-based RBF neural networks results in a structurally optimized structure and comes with a higher level of flexibility than the one encountered in the conventional RBF neural networks. The PSO-based design procedure being applied at each node of RBF neural networks leads to the selection of preferred parameters with specific local characteristics (such as the number of input variables, a specific set of input variables, and the distribution constant value in activation function) available within the RBF neural networks. To evaluate the performance of the proposed data-centroid RBF neural network with extended polynomial function, the model is experimented with using the nonlinear process data(2-Dimensional synthetic data and Mackey-Glass time series process data) and the Machine Learning dataset(NOx emission process data in gas turbine plant, Automobile Miles per Gallon(MPG) data, and Boston housing data). For the characteristic analysis of the given entire dataset with non-linearity as well as the efficient construction and evaluation of the dynamic network model, the partition of the given entire dataset distinguishes between two cases of Division I(training dataset and testing dataset) and Division II(training dataset, validation dataset, and testing dataset). A comparative analysis shows that the proposed RBF neural networks produces model with higher accuracy as well as more superb predictive capability than other intelligent models presented previously.

• Proceedings of the IEEK Conference
• /
• 2002.07a
• /
• pp.254-257
• /
• 2002

In previous study about combinatorial optimization problem solver by using neural network, since Hopfield method, to converge into the optimum solution sooner and certainer is regarded as important. Namely, only static states are considered as the information. However, from a biological point of view, the dynamical system has lately attracted attention. Then we propose the "dynamical" combinatorial optimization problem solver using hysteresis neural network. In this article, the proposal system is evaluated by the N-Queen problem.

• Proceedings of the Korea Inteligent Information System Society Conference
• /
• 2000.04a
• /
• pp.347-356
• /
• 2000

This paper compares four models of artificial neural networks (ANN) supported by genetic algorithms the prediction of stock price index. Previous research proposed many hybrid models of ANN and genetic algorithms(GA) in order to train the network, to select the feature subsets, and to optimize the network topologies. Most these studies, however, only used GA to improve a part of architectural factors of ANN. In this paper, GA simultaneously optimized multiple factors of ANN. Experimental results show that GA approach to simultaneous optimization for ANN (SOGANN3) outperforms the other approaches.

• Proceedings of the Korea Institute of Convergence Signal Processing
• /
• 2003.06a
• /
• pp.94-97
• /
• 2003

In this Paper, propose for MLP(multilayer perception) neural network that uses optimization recognition training scheme for the wavelet transform and the numeral image add to noise, and apply this system in Numeral Recognition. As important part of original image information preserves maximum using the wavelet transform, node number of neural network and the loaming convergence time did size of input vector so that decrease. Apply in training vector, examine about change of the recognition rate as optimization recognition training scheme raises noise of data gradually. We used original image and original image added 0, 10, 20, 30, 40, 50㏈ noise (or the increase of numeral recognition rate. In case of test image added 30∼50㏈, numeral recognition rate between the original image and image added noise for training Is a little But, in case of test image added 0∼20㏈ noise, the image added 0, 10, 20, 30, 40 , 50㏈ noise is used training. Then numeral recognition rate improved 9 percent.

• Proceedings of the KIEE Conference
• /
• 2005.05a
• /
• pp.3-5
• /
• 2005

In this paper. we introduce a new structure of fuzzy-neural networks Fuzzy Set-based Polynomial Neural Networks (FSPNN). The two underlying design mechanisms of such networks involve genetic optimization and information granulation. The resulting constructs are Fuzzy Polynomial Neural Networks (FPNN) with fuzzy set-based polynomial neurons (FSPNs) regarded as their generic processing elements. First, we introduce a comprehensive design methodology (viz. a genetic optimization using Genetic Algorithms) to determine the optimal structure of the FSPNNs. This methodology hinges on the extended Group Method of Data Handling (GMDH) and fuzzy set-based rules. It concerns FSPNN-related parameters such as the number of input variables, the order of the polynomial, the number of membership functions, and a collection of a specific subset of input variables realized through the mechanism of genetic optimization. Second, the fuzzy rules used in the networks exploit the notion of information granules defined over systems variables and formed through the process of information granulation. This granulation is realized with the aid of the hard C-Means clustering (HCM). The performance of the network is quantified through experimentation in which we use a number of modeling benchmarks already experimented with in the realm of fuzzy or neurofuzzy modeling.

• Nuclear Engineering and Technology
• /
• v.56 no.5
• /
• pp.1672-1678
• /
• 2024

We introduce a creative approach combining machine learning with optimization techniques to enhance the optimization of the loading pattern (LP). Finding the optimal LP is a critical decision that impacts both the reload safety and the economic feasibility of the nuclear fuel cycle. While simulated annealing (SA) is a widely accepted technique to solve the LP optimization problem, it suffers from the drawback of high computational cost since LP optimization requires three-dimensional depletion calculations. In this note, we introduce a technique to tackle this issue by leveraging neural networks to filter out inappropriate patterns, thereby reducing the number of SA evaluations. We demonstrate the efficacy of our novel approach by constructing a machine learning-based optimization model for the LP data of the Korea Standard Nuclear Power Plant (OPR-1000).

• 제어로봇시스템학회:학술대회논문집
• /
• 1993.10b
• /
• pp.76-80
• /
• 1993

In this paper we discuss optimization of neural network parameters, such as inclination of the sigmoid function, the numbers of the input layer's units and the hidden layer's units, considering application to recognition of hand written Hangul. Hangul characters are composed of vowels and consonants, and basically classified to six patterns by their positions. Using these characteristics of Hangul, the pattern of a given character is determined by its peripheral distribution and the other features. After then, the vowels and the consonants are recognized by the optimized neural network. The constructed recognition system including a neural network is applied to non-learning Hangul written by some Korean people, which are the names randomly taken from Korean spiritual and cultural research institute.

• Proceedings of the Korean Society of Machine Tool Engineers Conference
• /
• 2004.04a
• /
• pp.3-8
• /
• 2004

When most manufacturing company produce grinding product operators decide grinding condition by experience and subjective judgment. The study on grinding manufacture have been developed to get the grinding condition with the same result when non-experienced or experienced worker work. The objective of this study is to develope the grinding condition and predict the result of grinding by neural network. Several discussions were made in following areas as; getting MRR with image processing, the architecture optimization of neural network with experiment design, analysis of the input neurons using sensitivity approach. The results showed that the developed approach was the best method in predicting grinding condition with respect to surface finish quality.

• The Transactions of the Korean Institute of Electrical Engineers P
• /
• v.54 no.1
• /
• pp.20-29
• /
• 2005

This paper presents an implementation of efficiency optimization of reluctance synchronous motor (RSM) using a neural network (NN) with a direct torque control (DTC). The equipment circuit considered with iron losses in RSM is analyzed theoretically, and the optimal current ratio between torque current and exiting current component are derived analytically. For the RSM driver, torque dynamic can be maintained with DTC using TMS320F2812 DSP Controller even with controlling the flux level because a torque is directly proportional to the stator current unlike induction motor. In order to drive RSM at maximum efficiency and good dynamics response, the Backpropagation Neural Network is adapted. The experimental results are presented to validate the applicability of the proposed method. The developed control system show high efficiency and good dynamic response features with 1.0 [kW] RSM having 2.57 inductance ratio of d/q.

• Proceedings of the KIEE Conference
• /
• 2006.10c
• /
• pp.266-268
• /
• 2006

In this paper, we design a Linear Quadratic Gaussian(LQG) controller for active suspensions. We can improve the inherent suspension problem, trade-off between the ride quality and the suspension travel by selecting appropriate weights in the LQ-objective function. Using an optimization-algorithm, Evolution Strategy(ES), we find the proper control gains for selected frequencies, which have major effects on the vibrations of the vehicle's state variables. The frequencies and proper control gains are used for the neural network data. During a vehicle running, the trained on-line neural network is activated and provides the proper gains for non-trained frequencies.