• Proceedings of the Korean Society of Tribologists and Lubrication Engineers Conference
• /
• 2001.11a
• /
• pp.166-171
• /
• 2001

In order to improve the efficiency of the design process and the quality of the resulting design for the application-based exclusive rolling element bearings, this study propose design methodologies by using a genetic-based combinatorial optimization. By the presence of discrete variables such as the number of rolling element (standard component) and by the engineering point of views, the design problem of the rolling element bearing can be characterized by the combinatorial optimization problem as a fully discrete optimization. A genetic algorithm is used to efficiently find a set of the optimum discrete design values from the pre-defined variable sets. To effectively deal with the design constraints and the multi-objective problem, a ranking penalty method is suggested for constructing a fitness function in the genetic-based combinatorial optimization. To evaluate the proposed design method, a robust performance analyzer of ball bearing based on quasi-static analysis is developed and the computer program is applied to some design problems, 1) maximize fatigue life, 2) maximize stiffness, 3) maximize fatigue life and stiffness, of a angular contact ball bearing. Optimum design results are demonstrate the effectiveness of the design method suggested in this study. It believed that the proposed methodologies can be effectively applied to other multi-objective discrete optimization problems.

• Proceedings of the Korean Society for Noise and Vibration Engineering Conference
• /
• 2002.05a
• /
• pp.651-659
• /
• 2002

Component Mode Synthesis (CMS) is a dynamic substructuring technique to get an approximate eigensolutions of large degree-of-freedom structures divisible into several components. But, In practice. most of large structures are modeled by different teams of engineers. and their respective finite element models often require different mesh resolutions. As a result, the finite element substructure models can be non-conforming and/or incompatible. In this work, A hybrid version of component mode synthesis using a localized lagrange multiplier to treat the non-conforming mesh problem was derived. Evolution Strategies (ESs) is a stochastic numerical optimization technique and has shown a robust performance for solving deterministic problems. An ESs conducts its search by processing a population of solutions for an optimization problem based on principles from natural evolution. An optimization example for raising the first natural frequency of a plate structure using beam stiffeners was presented using hybrid component mode synthesis and robust evolution strategies (RES) optimization technique. In the example. the design variables are the positions and lengths of beam stiffeners.

• Transactions of the Korean Society for Noise and Vibration Engineering
• /
• v.18 no.12
• /
• pp.1217-1223
• /
• 2008

Vehicle vibration mostly originates from the road excitation and causes discomfort, fatigue and even injury to a driver. Vehicle ride comfort is one of the most important performance indices to achieve a high-quality vehicle design. Since design parameter variations inevitably result in the vehicle ride comfort variance, the variance characteristics should be analyzed in the early design stage of the vehicle. The vehicle ride comfort is often defined by an index which employs a weighted RMS value of the acceleration PSD of a seat position. The design solution is obtained through two steps in this study. An optimization problem to obtain a minimum ride comfort index is solved first. Then another optimization problem to obtain minimum variance of the ride comfort index is solved. For the optimization problems, the equations of motion and the sensitivity equations are derived basing on a 5-DOF vehicle model. The numerical results show that an optimal solution for the minimum ride comfort is not necessarily same as that of the minimum variance of the ride comfort.

• Korean Management Science Review
• /
• v.12 no.2
• /
• pp.105-127
• /
• 1995

This paper solves different machine-part group formation (MPGF) problems using genetic algorithms to demonstrate that it can be a new robust alternative to the conventional heuristic approaches for optimization problems. We first give an overview of genetic algorithms: Its principle, various considerations required for its implementation, and the method for setting up parameter values are explained. Then, we describe the MPGF problem which are critical to the successful operation of cellular manufacturing or flexible manufacturing systems. We concentrate on three models of the MPGF problems whose forms of the objective function and/or constraints are quite different from each other. Finally, numerical examples of each of the models descibed above are solved by using genetic algorithms. The result shows that the solutions derived by genetic algorithms are comparable to those obtained through problem-specific heuristic methods.

• Journal of the Korean Institute of Intelligent Systems
• /
• v.20 no.3
• /
• pp.375-379
• /
• 2010

The support vector machine (SVM) has been widely used in variety pattern recognition problems applicable to recommendation systems due to its strong theoretical foundation and excellent empirical successes. However, SVM is sensitive to the presence of outliers since outlier points can have the largest margin loss and play a critical role in determining the decision hyperplane. For robust SVM, we limit the maximum value of margin loss which includes the non-convex optimization problem. Therefore, we proposed the design method of robust SVM using genetic algorithm (GA) which can solve the non-convex optimization problem. To demonstrate the performance of the proposed method, we perform experiments on various databases selected in UCI repository.

• International Journal of Aeronautical and Space Sciences
• /
• v.4 no.1
• /
• pp.45-52
• /
• 2003

A three-dimensional aerodynamic shape optimization technique in inviscid compressible flows is developed by using a parallel continuous adjoint formulation on unstructured meshes. A new surface mesh modification method is proposed to overcome difficulties related to patch-level remeshing for unstructured meshes, and the effect of design sections on aerodynamic shape optimization is examined. Applications are made to three-dimensional wave drag minimization problems including an ONERA M6 wing and the EGLIN wing-pylon-store configuration. The results show that the present method is robust and highly efficient for the shape optimization of aerodynamic configurations, independent of the number of design variables used.

• International Journal of Control, Automation, and Systems
• /
• v.1 no.2
• /
• pp.184-193
• /
• 2003

Biaffine Matrix Inequality (BMI) is known to provide the most general framework in control synthesis, but problems involving BMI's are very difficult to solve because nonconvex optimization should be solved. In the previous paper, we proposed a new solver for problems involving BMI's using Evolutionary Algorithms (EA). In this paper, we solve several control synthesis examples such as Reduced-order control, Simultaneous stabilization, Multi-objective control, $H_{\infty}$ optimal control, Maxed $H_2$ / $H_{\infty}$control design, and Robust $H_{\infty}$ control. Each of these problems is formulated as the standard BMI form, and solved by the proposed algorithm. The performance in each case is compared with those of conventional methods.

• International Journal of Fuzzy Logic and Intelligent Systems
• /
• v.16 no.2
• /
• pp.81-86
• /
• 2016

Term weighting is a popular technique that effectively weighs the term features to improve accuracy in document classification. While several successful term weighting algorithms have been suggested, none of them appears to perform well consistently across different data domains. In this paper we propose several reasonable methods to combine different term weight vectors to yield a robust document classifier that performs consistently well on diverse datasets. Specifically we suggest two approaches: i) learning a single weight vector that lies in a convex hull of the base vectors while minimizing the class prediction loss, and ii) a mini-max classifier that aims for robustness of the individual weight vectors by minimizing the loss of the worst-performing strategy among the base vectors. We provide efficient solution methods for these optimization problems. The effectiveness and robustness of the proposed approaches are demonstrated on several benchmark document datasets, significantly outperforming the existing term weighting methods.

• Transactions of the Korean Society of Mechanical Engineers A
• /
• v.26 no.12
• /
• pp.2483-2491
• /
• 2002

This study presents an efficient method for robust optimal design. In order to avoid the excessive evaluations of the exact performance functions, two-point diagonal quadratic approximation method is employed for approximating them during optimization process. This approximation method is one of the two point approximation methods. Therefore, the second order sensitivity information of the approximated performance functions are calculated by an analytical method. As a result, this enables one to avoid the expensive evaluations of the exact $2^{nd}$ derivatives of the performance functions unlike the conventional robust optimal design methods based on the gradient information. Finally, in order to show the numerical performance of the proposed method, one mathematical problem and two mechanical design problems are solved and their results are compared with those of the conventional methods.

• Journal of the Korean Society of Manufacturing Process Engineers
• /
• v.8 no.4
• /
• pp.110-121
• /
• 2009

It is essential to embed product quality in the design process to win the global competition. Many components found in many products including automobiles and electronic devices are fabricated using sheet metal forming processes. Wrinkle and fracture are two types of defects frequently found in the sheet metal forming process. Reducing such defects is a hard problem as they are affected by many uncontrollable factors. Attempts to solve the problem based on traditional deterministic optimization theories are often led to failures. Furthermore, the wrinkle and fracture are conflicting defects in such a way that reducing one defect leads to increasing the other. Hence, it is a difficult task to reduce both of them at the same time. In this research, a new design method for reducing the rates of conflicting defects under uncontrollable factors is presented by using operating window and a sequential search procedure. A new SN ratio is proposed to overcome the problems of a traditional SN ratio used in the operating window technique. The method is applied to optimizing the robust design of a sheet metal forming process. To show the effectiveness of the proposed method, a comparison is made between the traditional and the proposed methods using simulation software, applied to a design of particular sheet metal forming process problem. The results show that the proposed method always gives a more robust design that is less sensitive to noises than the traditional method.