摘要:

The main objective of this paper is to develop procedures to formulate sensitivity-state equations for systems with multi-terminal first-order linear components, whether active or passive, without using equivalent circuit techniques. Thereby, the sensitivity-state variables that appear are all directly measureable, with evident advantages when systems are being studied in the laboratory as well as simulated on the digital computer. The equations are obtained through a well-documented algorithm for picking the maximally selected formulation forest of the system graph under consideration. Using the concepts of state space it is seen that sensitivity functions may be obtained by calculating the responses of a dependent linear model, topologically equivalent to the original, with element values and driving functions determined by partial derivatives of the original system quantities. The process of automatic formulation and solution of the two classes of state models for the original and sensitivity systems is greatly facilitated because of the basic structural similarity among the two. The procedure can be extended to cover higher-order linear multi-terminal components, non-linear systems and higher-order sensitivity functions.

ISSN:0020-7179    

DOI:10.1080/00207177108932033

出版商:Taylor & Francis Group    

年代:1971

数据来源: Taylor

摘要:

The quality of feedback laws of the type:acting on systems with unknown functions ƒ,xk+1: = ƒ(xk,uk), is studied in terms of the power which measures the convergence rate of the sequence {xk} → 0. Among a class of algorithms one with maximum power is found.

ISSN:0020-7179    

DOI:10.1080/00207177108932034

出版商:Taylor & Francis Group    

年代:1971

数据来源: Taylor

摘要:

Une méthode d'estimation simultanée du paramètre et de l'état d'un processus discret non stationnaire est exposée. Elle est basée sur une estimation linéaire et l'espérance mathématique conditionnelle. Les résultats de l'estimation sont utilisés pour une commande auto-adaptative avec un critère d'optimisation quadratique. La qualité de la méthode proposée est vérifiée pour deux exemples numériques.

ISSN:0020-7179    

DOI:10.1080/00207177108932035

出版商:Taylor & Francis Group    

年代:1971

数据来源: Taylor

摘要:

A partir de la méthode de l'estimation simultanée du paramètre et de l'état d'un processus discret du premier ordre, un processus généralisé est identifié par une estimation linéaire.

ISSN:0020-7179    

DOI:10.1080/00207177108932036

出版商:Taylor & Francis Group    

年代:1971

数据来源: Taylor

摘要:

A procedure is described for the design of stable adaptive model reference control systems. This procedure combines the Lyapunov method with the use of sensitivity coefficients. A basic algorithm is first obtained and the resulting system is shown to be asymptotically stable in the sense of ‘eventual stability’ as used by La Salle and Rath (1963). Modifications of this algorithm are then proposed which are easier to implement. One of these uses sensitivity coefficients and is studied in some detail.

ISSN:0020-7179    

DOI:10.1080/00207177108932037

出版商:Taylor & Francis Group    

年代:1971

数据来源: Taylor

摘要:

This paper describes a method of determining the stability at the sampling instants, of a discrete closed-loop system which includes a Kalman filter. This is done by determining thez-plane poles of a special augmented transition matrix. While this result stems from stochastic optimal control work, the method applies to any, not just optimal, values of the feedback matrix,M, and filter feedforward matrixK. It is still applicable when the filter model and the plant are dissimilar in coefficient values and order.

ISSN:0020-7179    

DOI:10.1080/00207177108932038

出版商:Taylor & Francis Group    

年代:1971

数据来源: Taylor

摘要:

The jump phenomena observed on the output signal of non-linear control systems subjected to non-stationary gaussian random inputs are presented. The analytical procedure is carried out by using the statistical linearization technique.

ISSN:0020-7179    

DOI:10.1080/00207177108932039

出版商:Taylor & Francis Group    

年代:1971

数据来源: Taylor

摘要:

An algorithm based on the potential function and the characteristic matrix is presented for multi-class pattern classification. The convergence of the algorithm is proved and a numerical example is given. It is shown that the algorithm using the Kesler transformation is a special case of the characteristic matrix potential function method.

ISSN:0020-7179    

DOI:10.1080/00207177108932040

出版商:Taylor & Francis Group    

年代:1971

数据来源: Taylor

摘要:

During the past five years, a number of papers have been published describing the use of specific optimal techniques in control and estimation. In this paper, many of those efforts are summarized and discussed. In addition, a generalized formulation of the specific optimal problem is presented which contains most of the preceding work as special cases.

ISSN:0020-7179    

DOI:10.1080/00207177108932041

出版商:Taylor & Francis Group    

年代:1971

数据来源: Taylor

摘要:

A method is presented whereby an optimal control may be obtained for a linear time-varying neutral system. The performance measure is quadratic with fixed finite upper limit. A form is derived for the coat functional, and the optimal control is obtained as the solution to a set of differential equations. A numerical technique for the solution of the equations is given and existence and uniqueness are proven.

ISSN:0020-7179    

DOI:10.1080/00207177108932042

出版商:Taylor & Francis Group    

年代:1971

数据来源: Taylor