• Proceedings of the Korean Institute of Intelligent Systems Conference
• /
• 2004.04a
• /
• pp.563-566
• /
• 2004

An intelligent Kalman filter (IKF) is proposed for tracking an incoming anti-ship missile. In the proposed IKF, the unknown target acceleration is regarded as an additive process noise. When the target maneuver is occurred, the residual of the Kalman filter increases in proportion to its magnitude. From this fact, the overall process noise variance can be approximated from the filter residual and its variation at every sampling time. A fuzzy system is utilized to approximate this valiance, and the genetic algorithm (GA) is applied to optimize the fuzzy system. In computer simulations, the tracking performance of the proposed IKF is compared with those of conventional maneuvering target tracking methods.

• Proceedings of the KIEE Conference
• /
• 2005.10b
• /
• pp.496-498
• /
• 2005

This paper presents intelligent digital redesign method of global approach for hybrid state space fuzzy-model-based controllers. For effectiveness and stabilization of continuous-time uncertain nonlinear systems under discrete-time controller, Takagi-Sugeno(TS) fuzzy model is used to represent the complex system. And global approach design problems viewed as a convex optimization problem that we minimize the error of the norm bounds between nonlinearly interpolated linear operators to be matched. Also by using the power series, we analyzed nonlinear system's uncertain parts more precisely. When a sampling period is sufficiently small, the conversion of a continuous-time structured uncertain nonlinear system to an equivalent discrete-time system have proper reason. Sufficiently conditions for the global state-matching of the digitally controlled system are formulated in terms of linear matrix inequalities (LMIs).

• Proceedings of the KIEE Conference
• /
• 2008.07a
• /
• pp.1545-1546
• /
• 2008

In this study, an image processing method with FRANSAC(Fuzzy RANSAC) is presented and discussed for the development of a microscopic gap measuring algorithm. Many problems in edge detection processing are mainly occurred by the illumination system. A serious problem is that the edge set of gap could include the error elements that have relatively larger error than normal. This problem leads to a incorrect measurement of gap. We present a gap measuring algorithm using FRANSAC[1] that is a representative robust estimation algorithm. FRANSAC is peformed by first categorizing all data into good sample set, bad sample set and vague sample set using a fuzzy classification and then sampling in only good sample set. Experimental results show that the presented gap measuring algorithm gives a higher accurate value of gap especially for the more noisy image data.

• Proceedings of the Korean Operations and Management Science Society Conference
• /
• 2002.05a
• /
• pp.877-884
• /
• 2002

This paper deals with a multi-stage TSK fuzzy modeling method by using Genetic Programming (GP). Based on the time sequence of sampling data the best structural change points of complex systems are detemined by using GP, and also the moving window is simultaneously introduced to overcome the excessive amount of calculation during the generating procedure of GP tree. Therefore, a multi-stage TSK fuzzy model that attempts to represent a complex problem by decomposing it into multi-stage sub-problems is addressed and its learning algorithm is proposed based on the Radial Basis Function (RBF) network. This approach allows us to determine the model structure and parameters by stages so that the problems ran be simplified.

• Proceedings of the KIEE Conference
• /
• 2007.07a
• /
• pp.1638-1639
• /
• 2007

This paper proposes an adaptive Dynamic Matrix Control (DMC) using Fuzzy Inference and its application to boiler-turbine system. Nine Step Response Models (SRM) at various operating points are represented as fuzzy inference rules. On-line fuzzy inference is performed at every sampling step to find the suitable SRM. Therefore, the proposed adaptive DMC can consider the nonlinearity of boiler-turbine system.

• Proceedings of the KIEE Conference
• /
• 1995.07a
• /
• pp.331-333
• /
• 1995

A high performance motor drive system must have a good speed command tracking, a insensitivity to a parameter variation and sampling time. In this paper, a robust speed controller for an induction motor is proposed. The speed controller is fuzzy-sliding adaptive controller and its system continuously is varied. That is, only P gain act in dynamic state, I gain in steady-state. Because this system is a sort of adaptive control system, global stability analysis is used to Lyapunov function. Consequently, in this paper application of fuzzy sliding adaptive controller to induction motor controlled by vecter control is presented and the control system is digitally implemented within DSP.

• Proceedings of the Korean Institute of Intelligent Systems Conference
• /
• 2005.11a
• /
• pp.91-94
• /
• 2005

This paper presents a new linear- matrix- inequality- basedintelligent digital redesign (LMI-based IDR) technique to match he states of the analog and the digital control systems at the intersampling instants as well as the sampling ones. The main features of the proposed technique are: 1) the multirate control is employed, and the control input is changed N times during one sampling period; 2) The proposed IDR technique is based on the compensated bilinear transformation.

• Journal of the Korean Institute of Intelligent Systems
• /
• v.17 no.3
• /
• pp.355-367
• /
• 2007

This paper proposes a design method of the PI(Proportional-Integral)+D(Derivative) controller using self-tuning of the design variables and controller gains. The used fuzzy PI+D controller is the approximated conventional continuos time linear PI+D controller and the used fuzzification method is the fuzzy single tone and the adapted defuzzification method is the simplified tenter of gravity. Fuzzy estimation result would be calculated in the other function elements from the classified fuzzy variables and the result determined by the design variables decides the controller gains. As a result, the proposed method shows the capability of the high speed tuning and can be applied to the case of input variables with many fuzzy partitions and also can bring out the advantage to reduce the reconstruction(digital sampling reconstruction) error. Most simulation results show that this controller makes much bettor efficiency and improvement by using design variables and controller gains.

• The KIPS Transactions:PartB
• /
• v.13B no.7 s.110
• /
• pp.671-678
• /
• 2006

In this paper, we propose a method for robustly estimating camera motion using fuzzy classification from the correspondences between two images. We use a RANSAC(Random Sample Consensus) algorithm to obtain accurate camera motion estimates in the presence of outliers. The drawback of RANSAC is that its performance depends on a prior knowledge of the outlier ratio. To resolve this problem the proposed method classifies samples into three classes(good sample set, bad sample set and vague sample set) using fuzzy classification. It then improves classification accuracy omitting outliers by iteratively sampling in only good sample set. The experimental results show that the proposed approach is very effective for computing a homography.

• Journal of Environmental Science International
• /
• v.15 no.2
• /
• pp.141-155
• /
• 2006

This study is aimed at the development of a runoff forecasting model to solve the uncertainties occurring in the process of rainfall-runoff modeling and improve the modeling accuracy of the stream runoff forecasting, The study area is the downstream of Naeseung-chun. Therefore, time-dependent data was obtained from the Wolpo water level gauging station. 11 and 2 out of total 13 flood events were selected for the training and testing set of model. The model performance was improved as the measuring time interval$(T_m)$ was smaller than the sampling time interval$(T_s)$. The Neuro-Fuzzy(NF) and TANK models can give more accurate runoff forecasts up to 4 hours ahead than the Feed Forward Multilayer Neural Network(FFNN) model in standard above the Determination coefficient$(R^2)$ 0.7.