• Journal of Institute of Control, Robotics and Systems
• /
• v.17 no.8
• /
• pp.777-782
• /
• 2011

Neural network technique is widely employed in the fields of signal processing, control systems, pattern recognition, etc. Learning of neural networks is an important procedure to accomplish dynamic system modeling. This paper presents a novel learning approach for differential neural network models based on the Kalman-Bucy filter theory. We construct an augmented state vector including original neural state and parameter vectors and derive a state estimation rule avoiding gradient function terms which involve to the conventional neural learning methods such as a back-propagation approach. We carry out numerical simulation to evaluate the proposed learning approach in nonlinear system modeling. By comparing to the well-known back-propagation approach and Kalman-Bucy filtering, its superiority is additionally proved under stochastic system environments.

• Journal of Institute of Control, Robotics and Systems
• /
• v.14 no.8
• /
• pp.779-784
• /
• 2008

This paper presents Kalman filter based system modeling algorithm for autonomous robot systems. State of the robot system is measured using embedded sensor systems and then carried to a host computer via ubiquitous sensor network (USN). We settle a linear state-space motion equation for unknown robot system dynamics and modify a popular Kalman filter algorithm in deriving suitable parameter estimation mechanism. To represent time-delay nature due to network media in system modeling, we construct an augmented state-space model which is mainly composed of original state and estimated parameter vectors. We conduct real-time experiment to test our proposed estimation algorithm where speed state of the constructed robot is used as system observation.

• Proceedings of the KIPE Conference
• /
• 2014.07a
• /
• pp.542-543
• /
• 2014

In this paper, an estimation scheme based on an augmented sigma-point Kalman filter to estimate the state of charge (SOC) of the battery is presented, where the battery parameters of the series resistance ($R_o$), diffusion capacitance ($C_1$) and resistance ($R_1$) are also estimated through the recursive least squares (RLS) for better accuracy of the SOC. The effectiveness of the proposed method is verified by simulation results.

• Proceedings of the Earthquake Engineering Society of Korea Conference
• /
• 2001.04a
• /
• pp.180-189
• /
• 2001

In this study, recently developed unscented Kalman filter (UKF) technique is studied for identification of nonlinear structural dynamic systems as an alternative to the extended Kalman filter (EKF). The EKF, which was originally developed as a state estimator for nonlinear systems, has been frequently employed for parameter identification by introducing the state vector augmented with the unknown parameters to be identified. However, the EKF has several drawbacks such as biased estimations and erroneous estimations especially for highly nonlinear dynamic systems due to its crude linearization scheme. To overcome the weak points of the EKF, the UKF was recently developed as a state estimator. Numerical simulation studies have been carried out on nonlinear SDOF system and nonlinear MDOF system. The results from a series of numerical simulations indicate that the UKF is superior to the EKF in the system identification of nonlinear dynamic systems especially highly nonlinear systems.

• Proceedings of the Computational Structural Engineering Institute Conference
• /
• 2001.10a
• /
• pp.117-126
• /
• 2001

In this study, the recently developed unscented Kalman filter (UKF) technique is studied for identification of nonlinear structural dynamic systems as an alternative to the extended Kalman filter (EKF). The EKF, which was originally developed as a state estimator for nonlinear systems, has been frequently employed for parameter identification by introducing the state vector augmented with the unknown parameters to be identified. However, the EKF has several drawbacks such as biased estimations and erroneous estimations especially for highly nonlinear dynamic systems due to its crude linearization scheme. To overcome the weak points of the EKF, the UKF was recently developed as a state estimator. Numerical simulation studies have been carried out on nonlinear SDOF system and nonlinear MDOF system. The results from a series of numerical simulations indicate that the UKF is superior to the EKF in the system identification of nonlinear dynamic systems especially highly nonlinear systems.

• The Transactions of the Korean Institute of Electrical Engineers A
• /
• v.48 no.5
• /
• pp.560-570
• /
• 1999

This paper formulates the suboptimal robust Kalman filtering problem into two coupled Linear Matrix Inequality (LMI) problems by applying Lyapunov theory to the augmented system which is composed of the state equation in the uncertain linear system and the estimation error dynamics. This formulations not only provide the sufficient conditions for the existence of the desired filter, but also construct the suboptimal robust Kalman filter. The proposed filter can guarantee the optimized upper bound of the estimation error variance for uncertain systems with parametric uncertainties in both the state and measurement matrices. In addition, this paper shows how the problem of finding the minimizing solution subject to Quadratic Matrix Inequality (QMI), which cannot be easily transformed into LMI using the usual Schur complement formula, can be successfully modified into a generic LMI problem.

• 제어로봇시스템학회:학술대회논문집
• /
• 2000.10a
• /
• pp.357-357
• /
• 2000

In this study, we consider a multi-input multi-output styrene polymerization reactor system for which the monomer conversion and the weight average molecular weight are controlled by manipulating the jacket inlet temperature and the feed flow rate. The reactor system is identified by using a linear subspace identification method and then the output feedback model predictive controller is constructed on the basis of the identified model. Here we use the Best Linear Unbiased Estimation (BLUE) filter as a stochastic estimator instead of the Kalman filter. The BLUE filter observes the state successfully without any a priori information of initial states. In contrast to the Kalman filter, the BLUE filter eliminates the offset by observing the state of the augmented system regardless of a priori information of the initial state for an integral white noise augmented system. A BLUE filter has a finite impulse response (FIR) structure which utilizes finite measurements and inputs on the most recent time interval [i-N, i] in order to avoid long processing times.

• Journal of Institute of Control, Robotics and Systems
• /
• v.14 no.8
• /
• pp.792-799
• /
• 2008

Motion estimation is crucial in a remote control for its convenience or accuracy. Time delays, however, can occur in the problem because data communication is required through a network. In this paper, state estimation problem with uncertain time delayed measurements is addressed. In dynamic system with noise, after taking measurements, it often requires some time until that is available in the filter algorithm. Standard filters not considering this time delays cannot be used since the current measurement is related with a past state. These delayed measurements are solved with augmented extended Kalman filter, and the uncertainty of delayed time is also resolved based on an explicit formulation. The proposed method is analyzed and verified by simulations.

• Journal of the Society of Naval Architects of Korea
• /
• v.38 no.4
• /
• pp.30-38
• /
• 2001

The current is considered in the conventional manoeuvering equation. This equation is represented as the nonlinear state and measurement equations in which external forces and the direction and the velocity of current are augmented as that variables. The external forces are modeled as the third-order Gauss-Markov processes and the direction and the velocity of current are assumed to be constant. The augmented state variables are estimated with extended Kalman-Bucy filter and the fixed-interval smoother. While Hwang estimated motion state variables, hydrodynamic coefficients and the current variables simultaneously by using extended Kalman filter, external forces of surge, sway and yaw and the direction and the velocity of current are the only parameters to be estimated in the estimation-before-modeling method. The current variables are satisfactorily estimated in simulation process where the measurement noise is present.

• 제어로봇시스템학회:학술대회논문집
• /
• 1997.10a
• /
• pp.1477-1480
• /
• 1997

This paper is concerned with the design of a suboptimal robust Kalman filter using LMI approach for system models in the state space, which are subjected to parameter uncertainties in both the state and measurement atrices. Under the assumption that augmented system composed of the uncertain system and the state estimation error dynamics should be stable, a Lyapunov inequality is obtained. And from this inequaltiy, the filter design problem can be transformed to the gneric LMI problems i.e., linear objective minimization problem and generalized eigenvalue minimization problem. When applied to uncertain linear system modles, the proposed filter can provide the minimum upper bound of the estimation error variance for all admissible parameter uncertainties.