• Journal of Mechanical Science and Technology
• /
• v.16 no.2
• /
• pp.203-210
• /
• 2002

For effective response surface modeling during sequential approximate optimization (SAO), the normalized and the augmented D-optimality criteria are presented. The normalized D-optimality criterion uses the normalized Fisher information matrix by its diagonal terms in order to obtain a balance among the linear-order and higher-order terms. Then, it is augmented to directly include other experimental designs or the pre-sampled designs. This augmentation enables the trust region managed sequential approximate optimization to directly use the pre-sampled designs in the overlapped trust regions in constructing the new response surface models. In order to show the effectiveness of the normalized and the augmented D-optimality criteria, following two comparisons are performed. First, the information surface of the normalized D-optimal design is compared with those of the original D-optimal design. Second, a trust-region managed sequential approximate optimizer having three D-optimal designs is developed and three design problems are solved. These comparisons show that the normalized D-optimal design gives more rotatable designs than the original D-optimal design, and the augmented D-optimal design can reduce the number of analyses by 30% - 40% than the original D-optimal design.

• Journal of the Computational Structural Engineering Institute of Korea
• /
• v.27 no.5
• /
• pp.345-352
• /
• 2014

In this paper, the CAD data for the optimal shape design obtained by isogeometric shape optimization is directly used to fabricate the specimen by using 3D printer for the experimental validation. In a conventional finite element method, the geometric approximation inherent in the mesh leads to the accuracy issue in response analysis and design sensitivity analysis. Furthermore, in the finite element based shape optimization, subsequent communication with CAD description is required in the design optimization process, which results in the loss of optimal design information during the communication. Isogeometric analysis method employs the same NURBS basis functions and control points used in CAD systems, which enables to use exact geometrical properties like normal vector and curvature information in the response analysis and design sensitivity analysis procedure. Also, it vastly simplify the design modification of complex geometries without communicating with the CAD description of geometry during design optimization process. Therefore, the information of optimal design and material volume is exactly reflected to fabricate the specimen for experimental validation. Through the design optimization examples of elasticity problem, it is experimentally shown that the optimal design has higher stiffness than the initial design. Also, the experimental results match very well with the numerical results. Using a non-contact optical 3D deformation measuring system for strain distribution, it is shown that the stress concentration is significantly alleviated in the optimal design compared with the initial design.

• The Korean Journal of Applied Statistics
• /
• v.28 no.1
• /
• pp.93-102
• /
• 2015

The D-optimal design for the nonlinear model typically depends on the unknown parameters to be estimated. Therefore, it is strongly recommended in literature to use a sequential experimental design for estimating the parameters. In this paper two stage experimental design is discussed under many different circumstances including estimating parameters. The method is so universal to be applied to any mixture of objectives for any model including linear model. A hybrid approach is suggested to handle more than 2 objectives in two-stage experimental design. The design is discussed in approximate design framework.

• Journal of Institute of Control, Robotics and Systems
• /
• v.21 no.11
• /
• pp.1003-1007
• /
• 2015

Depth calibration is a procedure for finding the conversion function that maps disparity data from a depth-sensing camera to actual distance information. In this paper, we present an optimal depth calibration method for Kinect$^{TM}$ sensors based on an experimental design and convex optimization. The proposed method, which utilizes multiple measurements from only two points, suggests a simplified calibration procedure. The confidence ellipsoids obtained from a series of simulations confirm that a simpler procedure produces a more reliable calibration function.

• Journal of the Computational Structural Engineering Institute of Korea
• /
• v.26 no.4
• /
• pp.241-246
• /
• 2013

From the numerical results of density-based topology design optimization, a CAD geometric model is constructed and fabricated using 3D printer to experimentally validate the optimal design. In the process of topology design optimization, we often experience checkerboard phenomenon and complicated branches, which could result in the manufacturing difficulty of the obtained optimal design. Sensitivity filtering and morphology methods are used to resolve the aforementioned issues. Identical volume fraction is used in both numerical and experimental models for precise validation. Through the experimental comparison of stiffness in various designs including the optimal design, it turns out that the optimal design has the highest stiffness and the experimental result of compliance matches very well with the numerical one.

• Journal of the Korean Magnetics Society
• /
• v.27 no.2
• /
• pp.54-62
• /
• 2017

This paper presents a design of the Interior Permanent Magnet Synchronous Motor (IPMSM). an initial design process is accomplished by using the parametric design. In the design process, motor characteristics of parameters is computed by the d-q axis equivalent circuit model. Then, an optimal design process is accomplished by combination the experimental design and the response surface method. Finally, the design and analysis results are verified with experimental results.

• Journal of the Korean Statistical Society
• /
• v.16 no.2
• /
• pp.71-79
• /
• 1987

The linear and quadratic log contrast model with mixtures on the strictly positive simplex, $$ x_{q-1} = {(x_1, \cdots, x_q):\sum x_, = 1 and \delta \leq \frac{x_i}{x_j} \leq \frac{1}{\delta} for all i,j},$$ are considered. Using the invariance arguments, symmetric D-optimal designs are investigated. The class of symmetric D-optimal designs for the linear log contrasts model is given. Any D-optimal design for the quadratic log contrast model is shown to metric D-optimal designs for q=3 and 4 cases are given.

• Journal of Korean Society for Quality Management
• /
• v.28 no.2
• /
• pp.135-146
• /
• 2000

This study has adapted a genetic algorithm for an optimal design for the first time. The models using a simulation are the nonlinear models. Using an genetic algorithm in D-optimal, it is more efficient than previous algorithms to get an object function. Not like other algorithms, without any troublesome restrictions about the initial solution, not falling into a local optimal solution, it's the most suitable algorithm. Also if we use it without any adding experiments, we can use it to find optimal design of experimental condition efficiently.

• The Korean Journal of Applied Statistics
• /
• v.4 no.2
• /
• pp.195-207
• /
• 1991

We consider the development of simple regression-like diagnostics for assessing the sensitivity of an optimal design to deviations from the assumptions of constant error variance. This contains a review of Cook's local influence approach and an application of local influence aproach to D-optimal experimental design. The method is applied in a number of simple examples in Section 3. Conclusions and directions for further research follow in Section 4.

• The Korean Journal of Applied Statistics
• /
• v.13 no.2
• /
• pp.393-405
• /
• 2000

Box and Draper( 19(5) listed some properties of a design that should be considered in design selection. But it is impossible that one design criterion from optimal experimental design theory reflects many potential objectives of an experiment, because the theory was originally based on the underlying model and its strict assumption about the error structure. Therefore, when it is neces::;ary to implement multi-objective experimental design. it is common practice to balance out the several optimal design criteria so that each design criterion involved benefits in terms of its relative "high" efficiency. In this study, we proposed several composite design criteria taking the case of heteroscedastic model. WVhen the heteroscedasticity is present in the model. the well known equivalence theorem between 1)- and C-optimality no longer exists and furthermore their design characteristics are sometimes drastically different. We introduced three different design criteria for this purpose: constrained design, combined design, and minimax design criteria. While the first two methods do reflect the prior belief of experimenter, the last one does not take it into account. which is sometimes desirable. Also we extended this method to the case when there are uncertainties concerning the error structure in the model. A simple algorithm and concluslOn follow.On follow.