摘要:

Controllers which minimize the maximum value of the magnitude of the system error are investigated for deterministic plants with saturating control signal. In order to obtain a unique optimal controller the minmax criterion is modified, resulting in what are termed strongly minmax systems. Explicit switching functions are found for feedback controllers in some low-order eases. In a third-order case it is found that the optimal system (whether minmax or strongly minmax) is not asymptotically stable.

ISSN:0020-7179    

DOI:10.1080/00207177808922357

出版商:Taylor & Francis Group    

年代:1978

数据来源: Taylor

摘要:

Recently, Monopoli lias developed a scheme for designing a stable model reference adaptive control system without requiring plant output derivatives, introducing the concept of an augmented error signal. In this paper we propose modifying his scheme by using a lower order reference model, reducing the augmented error equation into a first-order form and adding a proportional term to the integral adaptive law. These modifications offer the following advantages : (1) implementation of the overall adaptive control system can be simplified ; (2) the required adaptive law can be simply derived without using the Meyer-Kalman-Yacubovich lemma ; and (3) response characteristics of the adaptive system can be greatly improved. Computer simulation results are included to demonstrate the effectiveness of our scheme.

ISSN:0020-7179    

DOI:10.1080/00207177808922358

出版商:Taylor & Francis Group    

年代:1978

数据来源: Taylor

摘要:

This paper is concerned with a detailed study of controllability for a broad class of discrete linear stochastic systems with multiplicative state-dependent and control-dependent noise. Comparatively analysed are two approaches to the problem : the first employs a stochastic Liapunov-like function, and the second uses the well-known Chebyshev probability bound. It is shown for discrete systems with multiplicative noise that the Chebyshev approach is more suitable in general systems application.

ISSN:0020-7179    

DOI:10.1080/00207177808922359

出版商:Taylor & Francis Group    

年代:1978

数据来源: Taylor

摘要:

Recursive solution algorithms for a general class of deterministic linear-quadratic two-player non-zero sum difference games have been developed. Nash solutions are investigated for three kinds of information structures including the open-loop case, the feedback case and the open-closed case.

ISSN:0020-7179    

DOI:10.1080/00207177808922360

出版商:Taylor & Francis Group    

年代:1978

数据来源: Taylor

摘要:

A procedure for designing an optimal observer for a linear continuous time invariant system is described in this paper. The parameters of the observer are chosen so as to assign its eigenvalues to desired locations and at the same time a measure of the deviation of the estimate of the observed state from its actual value is minimized, A numerical example is included to show that the estimate of the state obtained by the optimal observer is much closer to the actual state than the estimate obtained from another observer with the same eigenvalues as the optimal observer.

ISSN:0020-7179    

DOI:10.1080/00207177808922361

出版商:Taylor & Francis Group    

年代:1978

数据来源: Taylor

摘要:

A matrix which consists of block elements is established in the block Schwarz form via a linear transformation. The transformation matrix constructed by Chen and Chu is modified and extended for transforming the block companion form to the block Schwarz form. A sufficient condition is derived for determining the stability of a multivariable system whose characteristics are expressed by a matrix polynomial. The matrix polynomial may or may not be symmetric.

ISSN:0020-7179    

DOI:10.1080/00207177808922362

出版商:Taylor & Francis Group    

年代:1978

数据来源: Taylor

摘要:

The structure of linear, time-invariant, completely controllable and observable multivariable systems under the action of constant output feedback control laws is studied. A structure theorem is stated and proved. Following Dickinson (1976), the notion of covariant output feedback control laws is introduced. It is shown that these results can be used to investigate the problems concerning (i) our ability to alter some of the Popov (1972) invariants by employing constant output feedback control, and (ii) the asymptotic behaviour of the closed loop poles under the variation of the output feedback gain matrix.

ISSN:0020-7179    

DOI:10.1080/00207177808922363

出版商:Taylor & Francis Group    

年代:1978

数据来源: Taylor

摘要:

The application of stochastic automata to the parameter self-optimization problem is considered theoretically. It is shown that the learning behaviours of stochastic automata in this problem can be reduced to the learning behaviours of stochastic automata under the non-stationary random environments R(Cl(t,ω ), ⃛, C(tr, ω)) which were considered in Baba and Sawaragi (1975).

ISSN:0020-7179    

DOI:10.1080/00207177808922364

出版商:Taylor & Francis Group    

年代:1978

数据来源: Taylor

摘要:

The problem of measuring the second-order Volterra kernel of a non-linear system by cross-correlation using a pseudo-random ternary test signal is considered. It is shown that the kernel values may be determined by a shift-and-add procedure, applied to the diagonal measurements obtained by removing a bias from the cross-correlation function of the system output and the square of the system input. The complete information content of experimental results may be extracted by this procedure, but it is shown to be preferable to extract only uniformly distributed kernel values, and the pseudo-random ternary signals which yield the greatest number of such values are tabulated. Implementation is shown to be simpler than conventional methods, and amenable to the application of discrete Fourier transforms. An example is used to illustrate the method.

ISSN:0020-7179    

DOI:10.1080/00207177808922365

出版商:Taylor & Francis Group    

年代:1978

数据来源: Taylor

摘要:

The problem of estimating the state of a stationary Gauss-Markov sequence observed in uncorrelated Gaussian noise of constant, hut unknown, covarianeeRis considered. A non-informative prior density is assigned to the prior innovations covariancoM, which is unknown by virtue of the uncertainty inR.Thea posterioridensity ofMevolves as an inverted Wishart, censored to account for the fact thatMis bounded from below by a positive definite matrix. The resulting non-linear state estimator involves a canonical integral which can be approximated to yield an attractive parallel filtering structure. The structure can be used to approximate the maximum aposteriori(MAP) estimate of the innovations covarianee,

ISSN:0020-7179    

DOI:10.1080/00207177808922366

出版商:Taylor & Francis Group    

年代:1978

数据来源: Taylor