• Proceedings of the Korean Institute of Intelligent Systems Conference
• /
• 2000.05a
• /
• pp.134-137
• /
• 2000

Genetic algorithms(GAs) have been widely used as a method to solve optimization problems. This is because GAs have simple and elegant tools with reproduction, crossover, and mutation to rapidly discover good solutions for difficult high-dimensional problems. They, however, do not guarantee the convergence of global optima in GA-hard problems such as deceptive problems. Therefore we proposed a Schema Co-Evolutionary Algorithm(SCEA) and derived extended schema 76988theorem from it. Using co-evolution between the first population made up of the candidates of solution and the second population consisting of a set of schemata, the SCEA works better and converges on global optima more rapidly than GAs. In this paper, we show advantages and efficiency of the SCEA by applying it to some problems.

• International Journal of Aeronautical and Space Sciences
• /
• v.15 no.2
• /
• pp.123-137
• /
• 2014

The aviation gas turbine is composed of many expensive and highly precise parts and operated in high pressure and temperature gas. When breakdown or performance deterioration occurs due to the hostile environment and component degradation, it severely influences the aircraft operation. Recently to minimize this problem the third generation of predictive maintenance known as condition based maintenance has been developed. This method not only monitors the engine condition and diagnoses the engine faults but also gives proper maintenance advice. Therefore it can maximize the availability and minimize the maintenance cost. The advanced gas turbine health monitoring method is classified into model based diagnosis (such as observers, parity equations, parameter estimation and Gas Path Analysis (GPA)) and soft computing diagnosis (such as expert system, fuzzy logic, Neural Networks (NNs) and Genetic Algorithms (GA)). The overview shows an introduction, advantages, and disadvantages of each advanced engine health monitoring method. In addition, some practical gas turbine health monitoring application examples using the GPA methods and the artificial intelligent methods including fuzzy logic, NNs and GA developed by the author are presented.

• The Transactions of The Korean Institute of Electrical Engineers
• /
• v.61 no.2
• /
• pp.303-311
• /
• 2012

In this study, we introduce a dynamic process model as well as the design methodology of optimized fuzzy controller for its efficient application to vacuum production system to produce a semiconductor, solar module and display and so on. In a vacuum control field, PID control method is widely used from the viewpoint of simple structure and preferred performance. But, PID control method is very sensitive to the change of environment of control system as well as the change of control parameters. Therefore, it's difficult to get a preferred performance results from target system which has a complicated structure and lots of nonlinear factors. To solve such problem, we propose the design methodology of an optimized fuzzy PID controller through a following series of steps. First a dynamic characteristic of the target system is analyzed through a series of experiments. Second the process model is built up and its characteristic is compared with real process. Third, the optimized fuzzy PID controller is designed using genetic algorithms. Finally, the fuzzy controller is applied to target system and then its performance is compared with that of other conventional controllers(PID, PI, and Fuzzy PI controller). The performance of the proposed fuzzy controller is evaluated in terms of auto-tuned control parameters and output responses considered by ITAE index, overshoot, rise time and steady state time.

• Proceedings of the Korean Institute of Intelligent Systems Conference
• /
• 2005.11a
• /
• pp.483-487
• /
• 2005

We introduce a new architecture of hetero-hybridized feed-forward neural networks composed of fuzzy set-based polynomial neural networks (FSPNN) and polynomial neural networks (PM) that are based on a genetically optimized multi-layer perceptron and develop their comprehensive design methodology involving mechanisms of genetic optimization and Information Granulation. The construction of Information Granulation based HFSPNN (IG-HFSPNN) exploits fundamental technologies of Computational Intelligence(Cl), namely fuzzy sets, neural networks, and genetic algorithms(GAs) and Information Granulation. The architecture of the resulting genetically optimized Information Granulation based HFSPNN (namely IG-gHFSPNN) results from a synergistic usage of the hybrid system generated by combining new fuzzy set based polynomial neurons (FPNs)-based Fuzzy Neural Networks(PM) with polynomial neurons (PNs)-based Polynomial Neural Networks(PM). The design of the conventional genetically optimized HFPNN exploits the extended Group Method of Data Handling(GMDH) with some essential parameters of the network being tuned by using Genetie Algorithms throughout the overall development process. However, the new proposed IG-HFSPNN adopts a new method called as Information Granulation to deal with Information Granules which are included in the real system, and a new type of fuzzy polynomial neuron called as fuzzy set based polynomial neuron. The performance of the IG-gHFPNN is quantified through experimentation.

• 한국경영정보학회:학술대회논문집
• /
• 2007.06a
• /
• pp.608-613
• /
• 2007

Prediction of corporate bankruptcies has long been an important topic and has been studied extensively in the finance and management literature because it is an essential basis for the risk management of financial institutions. Recently, support vector machines (SVMs) are becoming popular as a tool for bankruptcy prediction because they use a risk function consisting of the empirical error and a regularized term which is derived from the structural risk minimization principle. In addition, they don't require huge training samples and have little possibility of overfitting. However. in order to Use SVM, a user should determine several factors such as the parameters ofa kernel function, appropriate feature subset, and proper instance subset by heuristics, which hinders accurate prediction results when using SVM In this study, we propose a novel hybrid SVM classifier with simultaneous optimization of feature subsets, instance subsets, and kernel parameters. This study introduces genetic algorithms (GAs) to optimize the feature selection, instance selection, and kernel parameters simultaneously. Our study applies the proposed model to the real-world case for bankruptcy prediction. Experimental results show that the prediction accuracy of conventional SVM may be improved significantly by using our model.

• Proceedings of the Korea Inteligent Information System Society Conference
• /
• 2006.06a
• /
• pp.167-175
• /
• 2006

Building accurate corporate bankruptcy prediction models has been one of the most important research issues in finance. Recently, support vector machines (SVMs) are popularly applied to bankruptcy prediction because of its many strong points. However, in order to use SVM, a modeler should determine several factors by heuristics, which hinders from obtaining accurate prediction results by using SVM. As a result, some researchers have tried to optimize these factors, especially the feature subset and kernel parameters of SVM But, there have been no studies that have attempted to determine appropriate instance subset of SVM, although it may improve the performance by eliminating distorted cases. Thus in the study, we propose the simultaneous optimization of the instance selection as well as the parameters of a kernel function of SVM by using genetic algorithms (GAs). Experimental results show that our model outperforms not only conventional SVM, but also prior approaches for optimizing SVM.

• Journal of Biosystems Engineering
• /
• v.38 no.2
• /
• pp.149-162
• /
• 2013

Purpose: In image segmentation via thresholding, Otsu and Kapur methods have been widely used because of their effectiveness and robustness. However, computational complexity of these methods grows exponentially as the number of thresholds increases due to the exhaustive search characteristics. Methods: Particle swarm optimization (PSO) and genetic algorithms (GAs) can accelerate the computation. Both methods, however, also have some drawbacks including slow convergence and ease of being trapped in a local optimum instead of a global optimum. To overcome these difficulties, we proposed two new multi-level thresholding methods based on Bacteria Foraging PSO (BFPSO) and real-coded GA algorithms for fast segmentation. Results: The results from BFPSO and real-coded GA methods were compared with each other and also compared with the results obtained from the Otsu and Kapur methods. Conclusions: The proposed methods were computationally efficient and showed the excellent accuracy and stability. Results of the proposed methods were demonstrated using four real images.

• International Journal of Fuzzy Logic and Intelligent Systems
• /
• v.10 no.1
• /
• pp.12-18
• /
• 2010

This paper proposes an adaptive robust fuzzy control scheme for path tracking of a wheeled mobile robot with uncertainties. The robot dynamics including the actuator dynamics is considered in this work. The presented controller is composed of a fuzzy basis function network (FBFN) to approximate an unknown nonlinear function of the robot complete dynamics, an adaptive robust input to overcome the uncertainties, and a stabilizing control input. Genetic algorithms are employed to optimize the fuzzy rules of FBFN. The stability and the convergence of the tracking errors are guaranteed using the Lyapunov stability theory. When the controller is designed, the different parameters for two actuator models in the dynamic equation are taken into account. The proposed control scheme does not require the accurate parameter values for the actuator parameters as well as the robot parameters. The validity and robustness of the proposed control scheme are demonstrated through computer simulations.

• Journal of the Korean Institute of Intelligent Systems
• /
• v.11 no.3
• /
• pp.215-222
• /
• 2001

Neural network based models of semiconductor manufacturing processes have been shown to offer advantages in both accuracy and generalization over traditional methods. However, model development is often complicated by the fact that back-propagation neural networks contain several adjustable parameters whose optimal values unknown during training. These include learning rate, momentum, training tolerance, and the number of hidden layer neurOnS. This paper presents an investigation of the use of genetic algorithms (GAs) to determine the optimal neural network parameters for the modeling of plasma-enhanced chemical vapor deposition (PECVD) of silicon dioxide films. To find an optimal parameter set for the neural network PECVD models, a performance index was defined and used in the GA objective function. This index was designed to account for network prediction error as well as training error, with a higher emphasis on reducing prediction error. The results of the genetic search were compared with the results of a similar search using the simplex algorithm.

• International Journal of Aeronautical and Space Sciences
• /
• v.9 no.1
• /
• pp.100-110
• /
• 2008

It is not easy to monitor and identify all engine faults and conditions using conventional fault detection approaches like the GPA (Gas Path Analysis) method due to the nature and complexity of the faults. This study therefore focuses on a model based diagnostic method using Neural Network algorithms proposed for fault detection on a turbo shaft engine (PW 206C) selected as the power plant for a tilt rotor type unmanned aerial vehicle (Smart UAV). The model based diagnosis should be performed by a precise performance model. However component maps for the performance model were not provided by the engine manufacturer. Therefore they were generated by a new component map generation method, namely hybrid method using system identification and genetic algorithms that identifies inversely component characteristics from limited performance deck data provided by the engine manufacturer. Performance simulations at different operating conditions were performed on the PW206C turbo shaft engine using SIMULINK. In order to train the proposed BPNN (Back Propagation Neural Network), performance data sets obtained from performance analysis results using various implanted component degradations were used. The trained NN system could reasonably detect the faulted components including the fault pattern and quantity of the study engine at various operating conditions.