• The Korean Journal of Applied Statistics
• /
• v.17 no.2
• /
• pp.327-336
• /
• 2004

Many researchers have used the robust variogram to reduce the effect of outliers in spatial data analysis. Recently it is known that estimating the variogram after replacing outliers is more efficient. In this paper, we suggest a new data cleaner for geostatistic data analysis and compare the efficiency of outlier cleaners.

• Journal of the Korean Society of Manufacturing Process Engineers
• /
• v.16 no.1
• /
• pp.102-111
• /
• 2017

An automobile ball joint is the element for connecting the control arm and the knuckle arm, allowing rotational motion. The ball joint consists of the stud, plug, socket, and seat. These components are assembled through the caulking process that consists of plugging and spinning. In the existing research, the pull-out strength and gap stiffness were calculated, but we did not consider the uncertainties due to the numerical analysis and production. In this study, the uncertainties of material property and tolerance are considered to predict the distributions of pull-out strength and gap stiffness. Also, pull-out strength and gap stiffness are predicted as the a distribution rather than one deterministic value. Furthermore, a robust design applying the Taguchi method is suggested.

• Journal of the Institute of Convergence Signal Processing
• /
• v.10 no.2
• /
• pp.146-150
• /
• 2009

An Intelligent active roll angle controller design algorithm is discussed. The detailed mathematical formulation and analysis are discussed, and then modeling and design method for active roll angle controller are presented. This paper proposes a design method based upon intelligent robust controller design algorithm to control actively roll angle for improving cornering performance problems. The intelligent robust controller is designed for steady speed driving vehicle system model with representation of steering angle and yaw angular velocity parameters for cornering stability. And the detailed formulation and analysis for the objective vehicle system are investigated.

• The Journal of Korean Institute of Communications and Information Sciences
• /
• v.38A no.9
• /
• pp.798-809
• /
• 2013

In this paper, we propose robust relay and destination filter design methods for the multi-user peer-to-peer amplify-and-forward relaying systems while taking imperfect channel knowledge into consideration. Specifically, the relay and destination filter sets are developed to minimize the sum mean-squared-error (MSE). We first present a robust joint optimum relay and destination filter calculation method with an iterative algorithm. Motivated by the need to reduce computational complexity of the iterative scheme, we then formulate a simplified sum MSE minimization problem using the relay filter decomposability, which lead to two robust sub-optimum non-iterative design methods. Finally, we propose robust modified destination filter design methods which require only local channel state information between relay node and a specific destination node. The analysis and simulation results verify that, compared with the optimum iterative method, the proposed non-iterative schemes suffer a marginal loss in performance while enjoying significantly improved implementation efficiencies. Also it is confirmed that the proposed robust filter design methods provide desired robustness in the presence of channel uncertainty.

• 제어로봇시스템학회:학술대회논문집
• /
• 1993.10b
• /
• pp.81-87
• /
• 1993

In this paper robust stability conditions are obtained for linear dynamical systems under structured nonlinear time-varying perturbations, using absolute stability theory and the concept of dissipative systems. The conditions are expressed in terms of solutions to linear matrix inequality(LMI). Based on this result, a synthesis methodology is developed for robust feedback controllers with worst-case H$_{2}$ perforrmance via convex optimization and LMI formulation.

• Journal of the Korean Data and Information Science Society
• /
• v.17 no.4
• /
• pp.1309-1317
• /
• 2006

The paper considers some Bayes estimators of the finite population mean with auxiliary information under priors which are scale mixtures of normal, and thus have tail heavier than that of the normal. The proposed estimators are quite robust in general. Numerical methods of finding Bayes estimators under these heavy tailed priors are given, and are illustrated with an actual example.

• Journal of the Korean Data and Information Science Society
• /
• v.7 no.2
• /
• pp.203-210
• /
• 1996

A linear neural unit with a modified anti-Hebbian learning rule has been shown to be able to optimally fit curves, surfaces, and hypersurfaces by adaptively extracting the minor component of the input data set. In this paper, we study how to use the robust version of this neural fitting method for linear regression analysis. Furthermore, we compare this method with other methods when data set is contaminated by outliers.

• Journal of the Korean Data and Information Science Society
• /
• v.14 no.3
• /
• pp.471-479
• /
• 2003

Moving averages are the most popular statistics in analyzing time-series data like stock indices. However, moving averages are quite sensitive to unusual observations. In other words, they are not robust against unusual observations. We introduce the moving averages in terms of an M-estimator, and show how we can take advantages of using the proposed moving averages in fitting the data more than usual moving averages.

• Transactions on Control, Automation and Systems Engineering
• /
• v.1 no.1
• /
• pp.32-43
• /
• 1999

Adaptive robust control scheme is introduced for flexible joint manipulator with nonlinearities and uncertainties. The system does not satisfy the matching condition due to insufficient actuators for each node. The control only relies on the assumption that the bound of uncertainty exists. Thus, the bounded value does not need to be known a prior. The control utilizes the update law by estimating the bound of the uncertainties. The control scheme uses the backstepping method and constructs a state transformation. Also, stability analysis is done for both transformed system and original system.

• Proceedings of the KSME Conference
• /
• 2000.04a
• /
• pp.517-522
• /
• 2000

When a real robot manipulator is mathematically modeled. uncertainties are not avoidable. The uncertainties are often nonlinear and time-varying. The uncertain factors collie from imperfect knowledge ok system parameters. payload change. friction. external disturbance. and etc. In this paper. we propose a class of robust hybrid controls of manipulators without knowing the exact stiffness and provide the stability analysis. Simulation results are provided to show the effectiveness of the algorithms.