摘要:

The paper presents a linear predictor control and a process-model control for multi-variable systems with multiple delays in inputs and outputs subject to unmeasurable disturbances. The proposed control systems, like the Smith predictor, are simple for single-input and single-output systems, and can yield zero steady–state error with desired transient responses to the disturbances.

ISSN:0020-7179    

DOI:10.1080/00207178408933145

出版商:Taylor & Francis Group    

年代:1984

数据来源: Taylor

摘要:

The paper deals with the sensitivity to switching-function variations of a time–optimal system which is controlling a positional mechanism. The model chosen has a second-order differential equation with non-linear and discontinuous right–hand side. Such a model covers a large class of motion resistances, in particular all types of friction. Differential inequality theory and the theory of discontinuous differential equations, and in particular a method of generalizedjTsolutions, are used to show that time-optimal switching-function variations cause the origin to become an unstable equilibrium point of the state-plane and the control system to generate limit–cycles. Some suggestions for applications are given, as is a practical example.

ISSN:0020-7179    

DOI:10.1080/00207178408933146

出版商:Taylor & Francis Group    

年代:1984

数据来源: Taylor

摘要:

The aim of this work is to control highly non-linear plant by employing adaptive control strategies. It is proposed to identify system structure before parameter estimation takes place, i.e. to identify non-linear forms in the system description in a generalized way before estimating the coefficients of the linear and non-linear components of the system. In order to keep the estimates of the non-linear elements as generalized as possible, a polynomial approximation was used as the non-linear model. tt was proposed that a multivariable quadratic form would provide a reasonable approximation to any non-linearity identified, thus affording a great reduction in computation time for the identification of precise forms and hence providing an on-line capability. A control law was devised which would accept estimates of the non-linear description of the system. The stability of the non-linear controller was investigated under the quadratic regime used for estimation. An extension of Liapunov's indirect method was devised in which the coefficients of the locally approximating linear equations are non-linear, thus the indirect method is extended to indicate the asymptotic stability of a single point in a general region of asymptotic stability. Various simulation studies are shown which use the technique to control models of stable and unstable non-linear systems.

ISSN:0020-7179    

DOI:10.1080/00207178408933147

出版商:Taylor & Francis Group    

年代:1984

数据来源: Taylor

摘要:

This article presents some sufficient conditions for the transfer matrix of a state-estimator type feedback system to be proper or improper. These conditions are needed in the discussion of the well–posedness of the feedback system.

ISSN:0020-7179    

DOI:10.1080/00207178408933148

出版商:Taylor & Francis Group    

年代:1984

数据来源: Taylor

摘要:

Using an orthogonal decomposition technique, a matrix pseudo – in version lemma is presented to derive recursive algorithms for computing the best approximate solution to the sequential estimation problem. A deadbeat convergence property of the algorithms is shown and is applied to the design of a deadbeat adaptive tracking system. Simulations illustrate perfect tracking of the system output to the periodic reference trajectory in finite steps.

ISSN:0020-7179    

DOI:10.1080/00207178408933149

出版商:Taylor & Francis Group    

年代:1984

数据来源: Taylor

摘要:

A recursive algorithm of stochastic approximation type is derived for the estimation of space-dependent parameters in a class of distributed parameter systems driven by random inputs. The algorithm uses noisy observations taken over at a finite number of discrete points located in the spatial domain. Stochastic convergence of the estimates is proven. Simulation studies illustrate the good convergence property of the algorithm. The algorithm is much more simple both in structure and implementation and less restrictive on system structure to apply, compared to the algorithm of Kubrusly and Curtain (1977)

ISSN:0020-7179    

DOI:10.1080/00207178408933150

出版商:Taylor & Francis Group    

年代:1984

数据来源: Taylor

摘要:

A new algorithm is presented for reduced-order modelling of linear discrete-time systems, using an output–error criterion. A closed form expression is developed for the gradient of the cost function with respect to the model parameters. A computationally efficient algorithm for computing this gradient is derived. A Fletcher — Powell optimization procedure utilizes the gradient vector to compute the reduced-order model parameters. A special initialization procedure is proposed, and the stability of the reduced – order system is monitored. The performance of the algorithm is illustrated by some numerical examples.

ISSN:0020-7179    

DOI:10.1080/00207178408933151

出版商:Taylor & Francis Group    

年代:1984

数据来源: Taylor

摘要:

The accuracy properties of a general pseudo-linear regression (PLR) method are examined. Both off-line and on-line algorithms are considered. Assuming the parameter estimates converge they will under weak conditions be asymptotically gaussian distributed. Expressions for the corresponding covariance matrices are given. It is shown that the asymptotic covariance matrix of the off-line PLR algorithm is bounded from above by the matrix corresponding to the on-line PLR algorithm and from below by that corresponding to the prediction error method. Some simple numerical illustrations of the theoretical results are also included.

ISSN:0020-7179    

DOI:10.1080/00207178408933152

出版商:Taylor & Francis Group    

年代:1984

数据来源: Taylor

摘要:

This paper illustrates how the use of z–transform describing functions mapped into the w–domain provides a readily understood acceptable approach to the design of digital compensators for sampled relay control systems. The shortcomings of using the sinusoidal input describing function are also highlighted

ISSN:0020-7179    

DOI:10.1080/00207178408933153

出版商:Taylor & Francis Group    

年代:1984

数据来源: Taylor

摘要:

Time-weighted quadratic performance indices are evaluated for discrete and sampled-data linear systems. An evaluation in simple form is presented for discrete systems using a direct method based on manipulating the matrix infinite series. The result obtained is applied to sampled-data systems in which the performance index is given by the integral of continuous-data errors. It is shown that the result for sampled-data systems leads to the result for continuous systems when the sampling period approaches zero.

ISSN:0020-7179    

DOI:10.1080/00207178408933154

出版商:Taylor & Francis Group    

年代:1984

数据来源: Taylor