• Journal of information and communication convergence engineering
• /
• v.8 no.4
• /
• pp.370-376
• /
• 2010

Ensemble learning is a method for improving the performance of classification and prediction algorithms. However, its performance can be degraded due to multicollinearity problem where multiple classifiers of an ensemble are highly correlated with. This paper proposes genetic algorithm-based optimization techniques of SVM ensemble to solve multicollinearity problem. Empirical results with bankruptcy prediction on Korea firms indicate that the proposed optimization techniques can improve the performance of SVM ensemble.

• Transactions of the Korean Society of Automotive Engineers
• /
• v.10 no.4
• /
• pp.216-222
• /
• 2002

Brake Moan is a laud and strong noise occurring at any vehicle speed over 2 mph as a low frequency in below 600Hz. In this study, we targeted to shift the unstable mode that causes the brake moan from the moats frequency range to sufficiently higher frequency range to avoid the moan phenomenon. We simulated the finite element model and found out the nodes in which the brake moan occurs the most and we regarded the boundary and its relationship between the brake pad and the rotor as a spring coefficient k. With the binary set of the spring coefficient k, we finally used genetic algorithm (GA) to get the optimal topology of the brake pad and its shape to avoid the brake moan. The final result remarkably shows that genetic algorithm can be used in topology optimization procedures requiring complex eigenvalue problems.

• Journal of Communications and Networks
• /
• v.15 no.5
• /
• pp.486-495
• /
• 2013

In this article, we suggest a two-dimensional genetic algorithm (GA) method that applies a cognitive radio (CR) decision engine which determines the optimal transmission parameters for multicarrier communication systems. Because a CR is capable of sensing the previous environmental communication information, CR decision engine plays the role of optimizing the individual transmission parameters. In order to obtain the allowable transmission power of multicarrier based CR system demands interference analysis a priori, for the sake of efficient optimization, a two-dimensionalGA structure is proposed in this paper which enhances the computational complexity. Combined with the fitness objective evaluation standard, we focus on two multi-objective optimization methods: The conventional GA applied with the multi-objective fitness approach and the non-dominated sorting GA with Pareto-optimal sorting fronts. After comparing the convergence performance of these algorithms, the transmission power of each subcarrier is proposed as non-interference emission with its optimal values in multicarrier based CR system.

• 한국전산유체공학회:학술대회논문집
• /
• 2008.03b
• /
• pp.413-414
• /
• 2008

Abstract should be in English. The leaning angle optimization of turbine blade using the genetic algorithm was conducted in this paper. The calculation CFD technique was based upon the Diagonalized Alternating Directional Implicit scheme(DADI) with algebraic turbulencemodeling. The leaning angle of VKI turbine blade was represented using B-spline curve. The control points are the design variable. Genetic algorithm was taken into account as an optimization tool. The objective was to minimize the total pressure loss. The optimized final geometry shows the better aerodynamic performance compared with the initial turbine blade.

• Proceeding of EDISON Challenge
• /
• 2015.03a
• /
• pp.500-509
• /
• 2015

This paper presents the single objective design optimization of plate-fin heat sink equipped with fan cooling system using Genetic Algorithm. The proper heat sink and fan model are selected based on the previous studies. And the thermal resistance of heat sinks and fan efficiency during operation are calculated according to specific design parameters. The objective function is combination of thermal resistance and fan efficiency which have been taken to measure the performance of the heat sink. And Decision making procedure is suggested considering life time of semiconductor and Fan Operating cost. And also Analytical Model used for optimization is validated by Fluent, Ansys 13.0 and this model give a quite reasonable and reliable design.

• The Transactions of the Korean Institute of Electrical Engineers A
• /
• v.49 no.3
• /
• pp.85-94
• /
• 2000

In this paper, a robust H$\infty$ optimal tuning problem of a structure-specified PSS is investigated for power systems with parameter variation and disturbance uncertainties. Genetic algorithm is employed for optimization method of PSS parameters. The objective function of the optimization problem is the H$\infty$-norm of a closed loop system. The constraint of the optimization problem are based on the stability of the controller, limits on the values of the parameters and the desired damping of the dominant oscillation mode. It is shown that the proposed H$\infty$ PSS tuned using genetic algorithm is more robust than conventional PSS.

• Journal of the Society of Naval Architects of Korea
• /
• v.31 no.3
• /
• pp.12-18
• /
• 1994

Optimization in the engineering design is to select the best of many possible design alternatives in a complex design space. In order to optimize, various optimization methods have been used. One major problem of traditional optimization methods is that they often result in local optima. Recently genetic algorithm based on the mechanics of natural selection and natural genetics is used in many application fields for optimization. Genetic algorithm is more powerful to local optima, but it requires more calculation time and has difficulties in finding exact optimum point in design variable with real data type generally. In this paper. hybrid method was developed by coupling genetic algorithm and traditional direct search method. The developed method finds out a region for global optimum using genetic algorithm, and is to search global optimum using direct search method based on results obtained from genetic algorithm. By using hybrid method, calculation time is reduced and search efficient for optimum point is increased.

• Transactions of the Korean Society of Mechanical Engineers B
• /
• v.21 no.3
• /
• pp.405-413
• /
• 1997

A thermal optimization of the chip arrangement in the PCB channel oriented vertically and cooled by natural convection has been studied. The objective of this study is to find the chip arrangement that minimizes the maximum temperature of the entire PCB channel. SIMPLER algorithm is employed in the analysis, and the genetic algorithm is used for the optimization. The results show that the chip with a maximum volumetric heat generation rate has to be located at the bottom of the channel, and chips with relatively high heat generation rates should not be close to each other, and small chip should not be located between the large chips.

• Journal of Ocean Engineering and Technology
• /
• v.23 no.6
• /
• pp.67-70
• /
• 2009

Nesting and cutting path optimization have a great effect on the improvement of productivity in many industries such as shipbuilding, automotive, clothing, and so on. However, few researches have been carried out for the optimization of a cutting path algorithm. This study proposed a new method for cutting optimization using gravity center of cutting pieces and a genetic algorithm. The proposed method was tested for a sample plate including many different shapes of cutting pieces and compared to 2 other conventional methods. The test results showed that the new method had the shortest cutting path and the best effectiveness among the 3 methods.

• Journal of Korean Institute of Industrial Engineers
• /
• v.25 no.3
• /
• pp.360-368
• /
• 1999

We consider the problem of scheduling n jobs with sequence-dependent processing times on a set of parallel-identical machines. The processing time of each job consists of a pure processing time and a sequence-dependent setup time. The objective is to maximize the total remaining machine available time which can be used for other tasks. For the problem, a hybrid genetic algorithm is proposed. The algorithm combines a genetic algorithm for global search and a heuristic for local optimization to improve the speed of evolution convergence. The genetic operators are developed such that parallel machines can be handled in an efficient and effective way. For local optimization, the adjacent pairwise interchange method is used. The proposed hybrid genetic algorithm is compared with two heuristics, the nearest setup time method and the maximum penalty method. Computational results for a series of randomly generated problems demonstrate that the proposed algorithm outperforms the two heuristics.