• 제어로봇시스템학회:학술대회논문집
• /
• 제어로봇시스템학회 1996년도 한국자동제어학술회의논문집(국내학술편); 포항공과대학교, 포항; 24-26 Oct. 1996
• /
• pp.1205-1208
• /
• 1996

This paper proposes balanced model reduction of non-minimum phase plant. The algorithm presented in this paper is to convert high-order non-minimum phase plant into low-oder minimum phase plant using balanced model reduction. Balanced model reduction requires the error bound that Hankel singular value produces. This algorithm shows the tolerance that admits the method of this paper.

• 한국방송∙미디어공학회:학술대회논문집
• /
• 한국방송공학회 1998년도 학술대회
• /
• pp.71-75
• /
• 1998

압축된 비디오 신호를 전송함에 있어 에러에 의해 영상이 손상되는 영향을 최소화하기 위하여 여러 가지 오류은닉 기법들이 제안되고 있다. 본 논문에서는 인터 프레임에서 오류가 발생하여 정보가 손실되었을 경우, 움직임벡터의 효율적인 복구를 통한 오류은닉 방식을 제안하였다. 이를 위해 기존의 블록오류은닉 방식이 갖고 있는 문제점들을 살펴보고 이 중 전체 합 문제(total sum problem)에 대해 해결책으로 제시 될 수 있는 Infinite Norm을 이용한 블록경계정합방식을 제안하였고 실험에 의해 그 성능을 검증하였다.

• 한국시뮬레이션학회:학술대회논문집
• /
• 한국시뮬레이션학회 2001년도 The Seoul International Simulation Conference
• /
• pp.255-259
• /
• 2001

A skin tissue system is considered, whose configuration is electrically modelled by a layer over the infinite half space. The problem is to estimate the layer\`s thickness and respective conductivity from the measured impedance between the electrodes placed on the skin surface. A simulation is taken place as the optimization problem which is to minimize the norm between the \"measured\" and the solution for the assumed parameters.arameters.

• 제어로봇시스템학회:학술대회논문집
• /
• 제어로봇시스템학회 2004년도 ICCAS
• /
• pp.460-465
• /
• 2004

In this paper, we present some properties on receding horizon $H_{\infty}$ control for nonlinear discrete-time systems. First, we propose the nonlinear inequality condition on the terminal cost for nonlinear discrete-time systems. Under this condition, noninceasing monotonicity of the saddle point value of the finite horizon dynamic game is shown to be guaranteed. We show that the derived condition on the terminal cost ensures the closed-loop internal stability. The proposed receding horizon $H_{\infty}$ control guarantees the infinite horizon $H_{\infty}$ norm bound of the closed-loop systems. Also, using this cost monotonicity condition, we can guarantee the asymptotic infinite horizon optimality of the receding horizon value function. With the additional condition, the global result and the input-to-state stable property of the receding horizon value function are also given. Finally, we derive the stability margin for the saddle point value based receding horizon controller. The proposed result has a larger stability region than the existing inverse optimality based results.

• Earthquakes and Structures
• /
• 제10권6호
• /
• pp.1253-1272
• /
• 2016

For a nonlinear control system, there are many uncertainties, such as the structural model, controlled parameters and external loads. Although the significant progress has been achieved on the robust control of nonlinear systems through some researches on this issue, there are still some limitations, for instance, the complicated solving process, weak conservatism of system, involuted structures and high order of controllers. In this study, the computational structural mechanics and optimal control theory are adopted to address above problems. The induced norm is the eigenvalue problem in structural mechanics, i.e., the elastic stable Euler critical force or eigenfrequency of structural system. The segment mixed energy is introduced with a precise integration and an extended Wittrick-Williams (W-W) induced norm calculation method. This is then incorporated in the market-based control (MBC) theory and combined with the force analogy method (FAM) to solve the MBC robust strategy (R-MBC) of nonlinear systems. Finally, a single-degree-of-freedom (SDOF) system and a 9-stories steel frame structure are analyzed. The results are compared with those calculated by the $H{\infty}$-robust (R-$H{\infty}$) algorithm, and show the induced norm leads to the infinite control output as soon as it reaches the critical value. The R-MBC strategy has a better control effect than the R-$H{\infty}$ algorithm and has the advantage of strong strain capacity and short online computation time. Thus, it can be applied to large complex structures.

• 대한전자공학회:학술대회논문집
• /
• 대한전자공학회 2003년도 하계종합학술대회 논문집 V
• /
• pp.2713-2716
• /
• 2003

This paper presents a dynamic manipulability analysis method of the limb moving in viscous fluid. The key idea of the presented method is that the boundary of joint velocity can be converted to the velocity-dependant dynamic manipulability polytope through the coriolis, centrifugal and drag terms in dynamic equation. The velocity-dependant dynamic manipulability polytope is added to the inertial and restoring force manipulability polytope to get overall manipulability polytope of the limb moving in the fluid Each of the torque and velocity bounds arc considered in the infinite norm sense in joint space, and the drag force of a limb moving in fluid viscous is modeled as a quadratic form An analysis example with proposed analysis scheme is presented to validate the method.

• Structural Engineering and Mechanics
• /
• 제2권3호
• /
• pp.245-256
• /
• 1994

A new axisymmetric crack model is proposed on the basis of p-version of the finite element method limited to theory of small scale yielding. To this end, axisymmetric stress element is formulated by integrals of Legendre polynomial which has hierarchical nature and orthogonality relationship. The virtual crack extension method has been adopted to calculate the stress intensity factors for 3-D axisymmetric cracked bodies where the potential energy change as a function of position along the crack front is calculated. The sensitivity with respect to the aspect ratio and Poisson locking has been tested to ascertain the robustness of p-version axisymmetric element. Also, the limit value that is an exact solution obtained by FEM when degree of freedom is infinite can be estimated using the extrapolation equation based on error prediction in energy norm. Numerical examples of thick-walled cylinder, axisymmetric crack in a round bar and internal part-thorough cracked pipes are tested with high precision.

• 제어로봇시스템학회:학술대회논문집
• /
• 제어로봇시스템학회 1995년도 Proceedings of the Korea Automation Control Conference, 10th (KACC); Seoul, Korea; 23-25 Oct. 1995
• /
• pp.424-427
• /
• 1995

In this paper, a control law based on the receding horizon concept which robustly stabilizes time-varying discrete linear systems, is proposed. A dynamic game problem minimizing the worst case performance, is adopted as an optimization problem which should be resolved at every current time. The objective of the proposed control law is to guarantee the closed loop stability and the infinite horizon $H^{\infty}$ norm bound. It is shown that the objective can be achieved by selecting the proper terminal weighting matrices which satisfy the inequality conditions proposed in this paper. An example is included to illustrate the results..

• Transactions on Control, Automation and Systems Engineering
• /
• 제2권4호
• /
• pp.255-261
• /
• 2000

This paper presents the results of the robust H(sub)$\infty$ FIR filtering for a class of nonlinear continuous time-varying systems subject to real norm-bounded parameter uncertainty and know Lipschitz nonlinearity under sampled measurements. We address the problem of designing filters, using sampled measurements, which guarantee a prescribed H(sub)$\infty$ performance in continuous time-varying context, irrespective of the parameter uncertainty and unknown initial states. The infinite horizon causal H(sub)$\infty$FIR filter are investigated using the finite moving horizon in terms of two Riccati equations with finite discrete jumps.

• 제어로봇시스템학회논문지
• /
• 제10권10호
• /
• pp.930-939
• /
• 2004

In this paper, both dynamic constraints and kinematic constraints are considered for the analysis of manipulability of robotic systems comprised of multiple cooperating arms. Given bounds on the torques of each Joint actuator for every robot, the purpose of this study is to drive the bounds of task-space acceleration of object carried by the system. Bounds on each joint torque, described as a polytope, is transformed to the task-space acceleration through matrices related with robot dynamics, robot kinematics, object dynamics, grasp conditions, and contact conditions. A series of mathematical manipulations including the procedure calculating minimum infinite-norm solution of linear equation is applied to get the reachable acceleration bounds from given actuator dynamic constrains. Several examples including two robot systems as well as three robot system are shown with the assumptions of complete-constraint contact model(or' very soft contact') and insufficient or proper degree of freedom robot.