1. |
On a new definition of strict system equivalence |
|
International Journal of Control,
Volume 27,
Issue 5,
1978,
Page 657-672
A. C. PUGH,
A. K. SHELTON,
Preview
|
PDF (407KB)
|
|
摘要:
A new definition of strict system equivalence is considered and some of its properties are established. The underlying equivalence transformation for polynomial matrices is also investigated and a fundamental result obtained.
ISSN:0020-7179
DOI:10.1080/00207177808922402
出版商:Taylor & Francis Group
年代:1978
数据来源: Taylor
|
2. |
Convergence of an adaptive filter algorithm |
|
International Journal of Control,
Volume 27,
Issue 5,
1978,
Page 673-693
LENNART LJUNG,
Preview
|
PDF (594KB)
|
|
摘要:
A state-space representation of a dynamical, stochastic system is given. A corresponding model, parametrized in a particular way, is considered and an algorithm for the estimation of its parameters is analysed. The class of estimation algorithms thus considered contains general output error methods and model reference methods applied to stochastic systems. It also contains adaptive filtering schemes and, e.g. the extended least squares method.
ISSN:0020-7179
DOI:10.1080/00207177808922403
出版商:Taylor & Francis Group
年代:1978
数据来源: Taylor
|
3. |
The effect of compensator restrictions on pole-shifting |
|
International Journal of Control,
Volume 27,
Issue 5,
1978,
Page 695-703
P. A. COOK,
Preview
|
PDF (312KB)
|
|
摘要:
This paper investigates the extent to which the dynamics of a system can be altered by feedback compensation, subject to restrictions on the properties of the compensator.
ISSN:0020-7179
DOI:10.1080/00207177808922404
出版商:Taylor & Francis Group
年代:1978
数据来源: Taylor
|
4. |
Gain margins and root locus asymptotic behaviour in multivariable design Part I. The properties of the Markov parameters and the use of high feedback gain |
|
International Journal of Control,
Volume 27,
Issue 5,
1978,
Page 705-724
B. KOUVARITAKIS,
Preview
|
PDF (771KB)
|
|
摘要:
From a root locus point of view the ability of a feedback system to tolerate high gains depends on the distribution of the finite zeros (FZ) and the infinite zeros (IZ). The asymptotic directions of the closed loop poles that tend to the IZ with increasing scalar gain are determined by the eigenproperties of a set of parameters called the projected Markov parameters (PMP). Furthermore the points of radiation of the asymptotes called the pivots may also be computed in terms of the elementary state space matrices. The dependence of the order of the IZ and therefore the distribution of the asymptotes on the null structure of the PMP is discussed first and subsequently a design technique which aims at the reduction of the order of the IZ as well as the placement of the corresponding pivots is proposed. The effectiveness of such a design technique in improving the gain margins of feedback systems is illustrated by means of a numerical example.
ISSN:0020-7179
DOI:10.1080/00207177808922405
出版商:Taylor & Francis Group
年代:1978
数据来源: Taylor
|
5. |
Gain margins and root locus asymptotic behaviour in multivariable design Part II. A critical appraisal of frequency response methods from a root locus point of view |
|
International Journal of Control,
Volume 27,
Issue 5,
1978,
Page 725-751
B. KOUVARITAKIS,
Preview
|
PDF (878KB)
|
|
摘要:
A prerequisite for a linear multivariable feedback: system to have good gain margins is that the set of finite zeros and the set of infinite zeros lie in the left half of the complex plane. The spatial arrangement of the asymptotes to branches of the root locus vitally depends on the order of the infinite zeros and this in turn depends on the null structure of a set of parameters called the: projected Markov parameters. The study of the geometric structure of these parameters therefore may be employed as the basis of a critical appraisal of frequency response techniques from a root locus point of view; the inverse Nyquist array (INA) and characteristic locus (CL) techniques are considered here. It is shown that attempts to achieve diagonal dominance, a feature necessary for the application of the INA, may on certain occasions lead to infinite zeros of higher order and thus may severely curtail the system gain margins. The alignment of the characteristic directions at high frequencies, used by the CL on the other hand for the reduction of the misalignment angles, is shown to have a beneficial effect on the structure of the projected Markov parameters. Insight gained by such an appraisal reinforces the belief that root locus and frequency response methods used in a complementary manner may lead to a hybrid design philosophy which is more flexible and effective than either of the two approaches applied separately.
ISSN:0020-7179
DOI:10.1080/00207177808922406
出版商:Taylor & Francis Group
年代:1978
数据来源: Taylor
|
6. |
Control strategies for large-scale power systems |
|
International Journal of Control,
Volume 27,
Issue 5,
1978,
Page 753-767
J. D. DELACOUR,
M. DARWISH,
J. FANTIN,
Preview
|
PDF (433KB)
|
|
摘要:
In this paper, two control strategies are presented for the study of the control of a large-scale electric power system. In the first, the control is obtained by decomposition of the system to what are called ‘ ∈-coupled ’ subsystems, and in the second, a control strategy using closed-loop hierarchical control techniques is introduced. Finally, these two approaches are applied to a system consisting of three plants.
ISSN:0020-7179
DOI:10.1080/00207177808922407
出版商:Taylor & Francis Group
年代:1978
数据来源: Taylor
|
7. |
Synthesis of non-linear time-varying feedback systems with significant parameter uncertainty |
|
International Journal of Control,
Volume 27,
Issue 5,
1978,
Page 769-782
U. SHAKED,
Preview
|
PDF (410KB)
|
|
摘要:
A synthesis procedure for non-linear time-varying systems with large parameter uncertainty is presented. It guarantees that the system output to a given set of inputs will stay inside a prespecified bounded envelope of time functions in spite of the uncertainty and the possible existence of disturbances.
ISSN:0020-7179
DOI:10.1080/00207177808922408
出版商:Taylor & Francis Group
年代:1978
数据来源: Taylor
|
8. |
A note on a convergence theorem for a random optimization method |
|
International Journal of Control,
Volume 27,
Issue 5,
1978,
Page 783-788
N. BABA,
Preview
|
PDF (135KB)
|
|
摘要:
The objective of this note is to investigate the convergence of the random optimization method proposed by Matyas (1965). In his paper he proved that this method ensures convergence in probability to the optimal state. However, there can be found questionable points in his proof. This paper points out these questionable points and gives the modified theorem concerning the convergence of the random optimization method.
ISSN:0020-7179
DOI:10.1080/00207177808922409
出版商:Taylor & Francis Group
年代:1978
数据来源: Taylor
|
9. |
Input and output regularity of composite systems |
|
International Journal of Control,
Volume 27,
Issue 5,
1978,
Page 789-800
EROL SEZER,
ÖZAY HÜSEYIN,
Preview
|
PDF (391KB)
|
|
摘要:
Necessary and sufficient conditions are given for the input and/or output regularity of series and feedback connections of systems represented by system matrices. It is shown that both the series and the feedback connections of two systems can be made regular for both input and output by using almost arbitrary feedback around the systems, provided that both systems are of least order. The same result is also proved for the aeries connection of more than two systems.
ISSN:0020-7179
DOI:10.1080/00207177808922410
出版商:Taylor & Francis Group
年代:1978
数据来源: Taylor
|
10. |
The asymptotic joint distribution of the estimated autoregressive coefficients |
|
International Journal of Control,
Volume 27,
Issue 5,
1978,
Page 801-807
H. TONG,
Preview
|
PDF (191KB)
|
|
摘要:
The asymptotic joint distribution of the estimated autoregressive coefficients of an autoregressive model, whose order isnotnecessarily known, are derived. Our result is an extension of the classical result due to Mann and Wald (1943) in which the order is assumed to be knowna priori.
ISSN:0020-7179
DOI:10.1080/00207177808922411
出版商:Taylor & Francis Group
年代:1978
数据来源: Taylor
|