Browsing by Author "LIN, YP"

Now showing 1 - 15 of 15

DECENTRALIZED H-INFINITY CONTROLLER-DESIGN FOR LARGE-SCALE SYSTEMS
(1994-01-01)
CHENG, CF
;
WANG, WJ
;
LIN, YP
;
電控工程研究所
;
Institute of Electrical and Control Engineering
The problem of disturbance attenuation for large-scale systems using decentralized control is considered. The design procedure developed here involves constructing H infinity local state feedback controls via the Riccati equation approach, and a complete solution to this problem is carried out merely by iteratively solving a set of local algebraic Riccati equations. It is shown that when all the states are available, the H infinity norm of the closed-loop transfer function over all linear constant decentralized control laws is not imrpoved by allowing feedback to be dynamic. With successively smaller values of the prespecified disturbance level, the local state feedback H infinity suboptimal control problem can be solved and the controller can be chosen to be static.
DECENTRALIZED ROBUST-CONTROL OF DECOMPOSED UNCERTAIN INTERCONNECTED SYSTEMS
(1993-12-01)
CHENG, CF
;
WANG, WJ
;
LIN, YP
;
電控工程研究所
;
Institute of Electrical and Control Engineering
In this paper, a new design technique of decentralized robust controllers for a class of uncertain interconnected system is provided. The interconnection matrices and the uncertainties of the system are decomposed into a matched and a mismatched portions. Since the matched portion can always be compensated by feedback controls, this new technique can provide local controllers to withstand strong interconnections and large uncertainties. The relations among local feedback gains, uncertainty bounds, and interconnection strength are established to aid the robust controller design. A practical system with simulation results is given to exemplify the implementation of the controller design technique.
ERROR ESTIMATION OF INS GROUND ALIGNMENT THROUGH OBSERVABILITY ANALYSIS
(1992-01-01)
JIANG, YF
;
LIN, YP
;
電控工程研究所
;
Institute of Electrical and Control Engineering
A systematic analysis of the observability of an inertial navigation system (INS) in ground alignment with Bar-Itzhack and Berman's error model is presented. It is shown that the unobservable states are separately contained in two decoupled subspaces. The constraints on the selection of unobservable states are discussed. An estimation algorithm which is derived fully from the horizontal velocity outputs for computing the misalignment angles is provided. It reveals that the azimuth error can be entirely estimated from the estimates of leveling error and leveling error rate, without using gyro output signals explicitly.
EXTENDED LQG SELF-TUNING CONTROLLER IN A MIXED BACKGROUND
(1992-05-01)
LEE, BK
;
LIN, YP
;
電控工程研究所
;
Institute of Electrical and Control Engineering
We consider the problem of constructing a LQG-based adaptive controller so that the control system can track a deterministic command signal and simultaneously regulate stochastic disturbances. A new measure for mixed signal in the feedback loop is first introduced. Then a generalized cost function is formulated and the optimal 'extended LQG controller' can be constructed by solving a diophantine equation. Based on the extended LQG design, a self-tuning controller can be constructed. System tracking performance under ergodic process noise is analysed in detail and better results are obtained. Finally, the optimal tracking performance is attained if the parameter estimates converge to their true values.
EXTENSION OF LINEAR QUADRATIC OPTIMAL-CONTROL THEORY FOR MIXED BACKGROUNDS
(1991-10-01)
LEE, BK
;
CHEN, BS
;
LIN, YP
;
電子工程學系及電子研究所
;
電控工程研究所
The problem of finding an optimal controller, while both deterministic and stochastic signals are present in the feedback system, is addressed. A general framework is developed to solve the above problem and to unify conventional frequency domain LQ and LQG optimal control theories. Both the system transient and steady-state behaviours are considered and the exact tracking of the arbitrary reference signal is guaranteed. The decomposition of a mixed signal into the deterministic and stochastic parts gives a chance to investigate the effect of the weighting factors, which were specified in the cost function, on the compromise between transient and steady-state system performance. The solution, which characterizes the structure of the optimal controller, results in a set of independent diophantine equations, instead of a set of coupled ones. Moreover, the number of equations can be reduced under weak conditions. Finally, both transient command tracking and steady-state noise rejection capabilities can be optimized simultaneously by using a two-parameter control scheme.
EXTENSION OF LINEAR QUADRATIC OPTIMAL-CONTROL THEORY FOR MIXED BACKGROUNDS
(1991-10-01)
LEE, BK
;
CHEN, BS
;
LIN, YP
;
電子工程學系及電子研究所
;
Department of Electronics Engineering and Institute of Electronics
The problem of finding an optimal controller, while both deterministic and stochastic signals are present in the feedback system, is addressed. A general framework is developed to solve the above problem and to unify conventional frequency domain LQ and LQG optimal control theories. Both the system transient and steady-state behaviours are considered and the exact tracking of the arbitrary reference signal is guaranteed. The decomposition of a mixed signal into the deterministic and stochastic parts gives a chance to investigate the effect of the weighting factors, which were specified in the cost function, on the compromise between transient and steady-state system performance. The solution, which characterizes the structure of the optimal controller, results in a set of independent diophantine equations, instead of a set of coupled ones. Moreover, the number of equations can be reduced under weak conditions. Finally, both transient command tracking and steady-state noise rejection capabilities can be optimized simultaneously by using a two-parameter control scheme.
IMPROVED STRAPDOWN CONING ALGORITHMS
(1992-04-01)
JIANG, YF
;
LIN, YP
;
電控工程研究所
;
Institute of Electrical and Control Engineering
Three improved algorithms for strapdown attitude computation utilizing accumulated gyro increments from the previous and current update are developed and evaluated analytically under classical coning motion. The error criterion of Miller [6] is now derived directly from the rotation vector concept. The accuracy of updating rotation vector estimation can be improved at least two orders of magnitude as compared with those of conventional algorithms. The proposed algorithm is equivalent to increasing the number of gyro samples used in the conventional method and it requires less computer loading also.
MODEL-REFERENCE DEADBEAT CONTROL
(1991-11-01)
LEE, BK
;
CHEN, BS
;
LIN, YP
;
電子工程學系及電子研究所
;
電控工程研究所
The performance of deadbeat control systems is addressed. Analysis of conventional deadbeat control (CDC) reveals that, in addition to deadbeat tracking performance, other system properties, such as robustness and noise immunity, may need improvement. A model reference deadbeat control (MRDC) is proposed to remedy such deficiencies of CDC systems. Several examples are given to show that the MRDC scheme is better than the CDC scheme in attaining good transient response, noise immunity and robustness if a suitable reference model is selected.
NECESSARY AND SUFFICIENT CONDITION FOR STABILITY ROBUSTNESS UNDER NONLINEAR TIME-VARYING UNCERTAINTIES
(1990-02-01)
CHOU, HH
;
CHEN, BS
;
LIN, YP
;
電子工程學系及電子研究所
;
電控工程研究所
OBSERVER SYNTHESIS OF NONLINEAR INTERCONNECTED SYSTEMS BY CONSTRAINED RICCATI APPROACH
(1995-09-01)
CHENG, CF
;
WANG, WJ
;
LIN, YP
;
電控工程研究所
;
Institute of Electrical and Control Engineering
This paper provides a method for observer synthesis of interconnected systems with nonlinearities in subsystems and interconnection terms, which satisfy the Lipschitz-type conditions instead of any matching conditions. A new observer scheme is proposed under a local observability assumption, and the respective observer gains are obtained through Riccati equations with linear constraints. Solvability conditions for such a Constrained Riccati Problem (CRP) are derived, and a systematic procedure for obtaining the solution of a constrained Riccati equation is presented. Furthermore, the relationship between tolerable Lipschitz constants of the modeling nonlinearities and observer gains is established.
ON THE ROBUSTNESS OF THE STABILITY OF LYAPUNOV-TYPE STATE-FEEDBACK SYSTEMS
(1993-09-01)
CHENG, CF
;
WANG, WJ
;
LIN, YP
;
電控工程研究所
;
Institute of Electrical and Control Engineering
A generalized Lyapunov-type state feedback design for uncertain systems is proposed. Based on the solution of the proposed Lyapunov-type equation, bounds of the nonlinear (linear), time-varying/invariant uncertainties are obtained to maintain the system's stability. No statistical description of uncertain elements is assumed, and it is not necessary for the so-called ''matching condition'' to be satisfied, either.
ROBUST OBSERVER SYNTHESIS FOR NONLINEAR LARGE-SCALE SYSTEMS
(1994-06-01)
CHENG, CF
;
WANG, WJ
;
LIN, YP
;
電控工程研究所
;
Institute of Electrical and Control Engineering
The observer synthesis of nonlinear large-scale systems is discussed. The non-linearities satisfy the Lipschitz conditions instead of any matching conditions and exist on the interconnection terms and local system matrices. With the aid of the solution of the Riccati-type equation, the relationship between the nonlinearity bounds and the observer gains is derived. Moreover, the existence condition and seeking algorithm for the solution of the Riccati equation are investigated. An example with simulations is given to illustrate the application of our results.
ROBUST OPTIMIZATION OF MULTIVARIABLE FEEDBACK-SYSTEMS WITH TIME-VARYING NONLINEAR UNCERTAINTIES
(1993-03-01)
TU, MH
;
LIN, CM
;
LIN, YP
;
電控工程研究所
;
Institute of Electrical and Control Engineering
A design criterion is developed to achieve the input-output decoupling of multivariable feedback systems and the robust stabilization of systems with time-varying nonlinear uncertainties. Moreover, an effective design algorithm is derived to achieve the robust optimization of multivariable feedback systems subjected to time-varying nonlnear uncertainties. The theory of minimum H(infinity)-norm and the optimal interpolation technique are employed to solve this robust optimization problem. Since the requirements of internal stability are satisfied, this design algorithm performs appropriately, even if the plant is unstable and/or non-minimum phase. From the result of the robust optimization, we can predict the maximum sector bounds of nonlinear uncertainties that can be tolerated in the multivariable feedback system.
ROBUST POLE PLACEMENT - A FREQUENCY-DOMAIN APPROACH
(1990-02-01)
CHOU, HH
;
CHEN, BS
;
LIN, YP
;
電子工程學系及電子研究所
;
電控工程研究所
SYNTHESIS OF MODEL-REFERENCE LQG OPTIMAL SYSTEMS
(1993)
LIN, YP
;
TU, MH
;
電控工程研究所
;
Institute of Electrical and Control Engineering