• 한국통계학회:학술대회논문집
• /
• 한국통계학회 2003년도 추계 학술발표회 논문집
• /
• pp.257-262
• /
• 2003

We introduce a new robust regression estimator, self-tuning regression estimator. Various robust estimators have been developed with discovery for theories and applications since Huber introduced M-estimator at 1960's. We start by announcing various robust estimators and their properties, including their advantages and disadvantages, and furthermore, new estimator overcomes drawbacks of other robust regression estimators, such as ineffective computation on preserving robustness properties.

• Journal of the Korean Data and Information Science Society
• /
• 제24권3호
• /
• pp.659-667
• /
• 2013

If the regression model or the propensity model is correct, the unbiasedness of the estimator using doubly robust imputation can be guaranteed. Using a neural network instead of a logistic regression model for the propensity model, the estimators using doubly robust imputation are approximately unbiased even though both assumed models fail. We also propose a doubly robust estimator of ratio form using population information of an auxiliary variable. We prove some properties of proposed theory by restricted simulations.

• 제어로봇시스템학회논문지
• /
• 제7권4호
• /
• pp.363-363
• /
• 2001

We study on the velocity matching algorithm for transfer alignment of inertial navigation system(INS) using a robust H₂ filter. We suggest an uncertainty model and a discrete robust H₂filter for INS and apply the suggested robust H₂ filter to the uncertainty model. The discrete robust H₂filter is shown by simulation to have better performance time and accuracy than Kalman filter.

• KIEE International Transactions on Power Engineering
• /
• 제4A권3호
• /
• pp.109-114
• /
• 2004

In this paper, the H$\infty$ control method is applied in the control of SPFC and generators in order to obtain the robust stability performance of the power system. Moreover, a SPFC H$\infty$ controller based on loop shaping is designed to achieve first-rate tracking properties in addition to the robust stability. This robust control method is applied to a Ward and Hale 6-bus power system and the effects are analyzed. The MATLAB program is used for simulation of the robust controller.

• 제어로봇시스템학회논문지
• /
• 제9권10호
• /
• pp.755-767
• /
• 2003

This paper proposes a robust fault-tolerant control framework for robot manipulators to maintain the required performance and achieve task completion in the presence of both partial joint failures and complete joint failures and uncertainties. In the case of a complete joint failure or free-swinging joint failure causing the complete loss of torque on a joint, a fully-actuated robot manipulator can be viewed as an underactuated robot manipulator. To detect and identify a complete actuator failure, an on-line fault detection operation is also presented. The proposed fault-tolerant control system contains a robust adaptive controller overcoming partial joint failures based on robust adaptive control methodology, an on-line fault detector detecting and identifying complete joint failures, and a robust adaptive controller overcoming partial and complete joint failures, and so eventually it can face and overcome joint failures and uncertainties. Numerical simulations are conducted to validate the proposed robust fault-tolerant control scheme.

• 대한기계학회논문집A
• /
• 제28권9호
• /
• pp.1368-1383
• /
• 2004

Robust design has been developed with the expectation that an insensitive design can be realized. That is, a product designed by robust design should be insensitive to external noises or tolerances. Robust design can be classified into three methods: (1) The Taguchi method (2) robust optimization (3) robust design by the axiomatic approach. In this paper, each method is reviewed and investigated. Pros and cons fur each method are discussed and a future direction for development is proposed.

• Journal of Mechanical Science and Technology
• /
• 제20권8호
• /
• pp.1169-1182
• /
• 2006

Robust design technology has been applied to versatile engineering problems to ensure consistency in product performance. Since 1980s, the concept of robust design has been introduced to numerical optimization field, which is called the robust optimization. The robustness in the robust optimization is determined by a measure of insensitiveness with respect to the variation of a response. However, there are significant difficulties associated with the calculation of variations represented as its mean and variance. To overcome the current limitation, this research presents an implementation of the approximate statistical moment method based on kriging metamodel. Two sampling methods are simultaneously utilized to obtain the sequential surrogate model of a response. The statistics such as mean and variance are obtained based on the reliable kriging model and the second-order statistical approximation method. Then, the simulated annealing algorithm of global optimization methods is adopted to find the global robust optimum. The mathematical problem and the two-bar design problem are investigated to show the validity of the proposed method.

• 한국안전학회지
• /
• 제27권1호
• /
• pp.76-85
• /
• 2012

This paper proposes a robust damper design technique for adjacent structures against model uncertainty. This approach introduces multi-objective optimization based system identification using measurement information which enables reasonable selection of the perturbation range in the robust design. Moreover, in order to improve the numerical efficiency in sampling the structural models required for the robust design of large structures, we define new objective functions which enable us to minimize the number of candidate models suitable to the purpose of the robust design. In addition, the performance index is newly employed to evaluate the robust performance of the sampled structural models, and the robust design has been performed according to the performance index. As a numerical example to demonstrate the efficiency of the proposed method, 5-story and 10-story two adjacent buildings are taken into account, and the existing and newly proposed robust design approaches are compared with each other. The results demonstrate that the proposed approach can guarantee more robust damper system only using small number of samples of the structural models because of using the measurement information which leads to improvement in the numerical efficiency, compared with the existing robust design methods.

• 한국기계가공학회지
• /
• 제7권3호
• /
• pp.3-11
• /
• 2008

Robust design pioneered by Dr. G. Taguchi has been applied to versatile engineering problems for improving quality. Since 1980s, the Taguchi method has been introduced to numerical optimization, complementing the deficiencies of deterministic optimization, which is often called the robust optimization. In this study, the robust optimization strategy is proposed by considering the robustness of objective and constraint functions. The statistics of responses in the functions are surrogated by kriging models. In addition, objective and/or constraint function is represented by the probability of success, thus facilitating robust optimization. The mathematical problem and the two-bar design problem are investigated to show the validity of the proposed method.

• 충청수학회지
• /
• 제26권4호
• /
• pp.723-734
• /
• 2013

A robust optimization problem, which has a maximum function of continuously differentiable functions as its objective function, continuously differentiable functions as its constraint functions and a geometric constraint, is considered. We prove a necessary optimality theorem and a sufficient optimality theorem for the robust optimization problem. We formulate a Wolfe type dual problem for the robust optimization problem, which has a differentiable Lagrangean function, and establish the weak duality theorem and the strong duality theorem which hold between the robust optimization problem and its Wolfe type dual problem. Moreover, saddle point theorems for the robust optimization problem are given under convexity assumptions.