G. C. Goodwin and R. L. Payne. Dynamic system identification: experiment design and data analysis. Academic Press, 1977.  Home/Search   Document Not in Database   Summary   Related Articles   Check

....required for estimation of the timing differences. Similarly, the performance analysis of the least squares estimator is studied to determine the tradeoff between the estimation error and the number of data symbols. The classical result of the least squares estimator focuses on the residual error [7]. This paper takes a different perspective by analyzing the error of the estimated parameter itself and exploits the well known results of the inverted Wishart function [8] 9] in the multivariate statistics area. A. Timing Difference Estimation The cross correlation technique has been used to ....

G. C. Goodwin and R. L. Payne, Dynamic System Identification: Experiment Design and Data Analysis. New York: Academic, 1977.

....x2.1 we will discuss in more detail the uncertainty structure that is used troughout the paper. Given two samples x s ; y s of the random vectors x; y, we seek an estimate x of x such that the worst case (with respect to the uncertainty Delta) a posteriori likelyhood of the samples is maximized, [14]. The likelyhood function L is defined as the logarithm of the a posteriori probability density L(x; Deltajx s ; y s ) log (f x (x s )f y (y s ) where f x ; f y are the probability density functions of x; y, respectively. Being x; y independent gaussian vectors, maximizing the likelyhood with ....

G.C. Goodwin and R.L. Payne, Dynamic System Identification: experiment design and data analysis. Academic Press, New York, 1977. 19

....to the uncertainty) a posteriori probability of the observed event. When no deterministic uncertainty is present on the model, this is the celebrated maximum likelihood (ML) approach to parameter estimation, which enjoys special properties such as efficiency and unbiasedness, see for instance [3, 15, 19]. For the important special case of linear estimation, we will discuss in Section 3.2 how these properties extend to the robust estimator, and how the resulting estimate is related to the minimum a posteriori variance estimator. To cast our problem in a ML setting, the log likelihood function L is ....

G.C. Goodwin and R.L. Payne, Dynamic System Identification: experiment design and data analysis. New York: Academic Press, 1977.

.... areavailable(e.g. 3 Hzsamplingfrequencysonarprofilers) theestimatedvelocitystandarddeviationislowerthan3 of thevelocity.Oncethedragparametershavebeendetermined,a suboptimalsinusoidalforce torqueinputisdesignedinorderto identifythevehicleinertia.Suchinputguaranteesobservability [37]andminimizestheturbulencegeneratednexttothepropellers. A.IdentificationoftheDragandThrusterInstallation Coefficients Understeady stateconditions,i.e. whentheapplied force torqueisconstant, 3)becomes (7) 3 wheretheunknownparametersarethelinearandquadraticdrag ....

....with rangingfrom1to arethethrusterinstallationcoefficients correspondingtothe differentthrustmappingmodes. AccordingtoLStheory,thestandarddeviation oftheestimatedparameter iscomputedas (11) where istheGaussianzeromeanmeasurementnoisevariance. Assuggestedin[37],ifsuchvarianceisunknown,itcan beestimatedby (12) Intheremainerofthepaper,thequantity willbe referredtoasthepercentileparametererror. B.IdentificationoftheInertiaCoefficients Havingidentifiedthedragparametersasdescribedabove,the ....

[Article contains additional citation context not shown here]

C.G.GoodwinandR.L.Payne,DynamicSystemIdentification:Exper- imentDesignandDataAnalysis,NewYork:Academic,1977.

....via the corresponding eigenvector V (i) Moreover, even in estimation problems with insufficient excitation, these non persistent contributions can easily dominate the noise energy as illustrated in Section 5. This is in contrast with classical excitations (i.e. pseudo random input signals [1, 2]) for which s 2 i can be expressed as Ns 2 0i with s 2 0i constant, for the Phi (i) V s are stationary over the data length N . 4.2 Eigen parameters : V (N) Assuming that e(t) is a Gaussian white noise (i.e. N(0; oe 2 ) and that the excitation assumption (18) holds, we can ....

G.C. Goodwin and R.L. Payne, Dynamic System Identification : Experiment Design and Data Analysis, Mathematics in Science and Engineering, Academic Press, Vol. 136, 1977, pp 124-207.

....least squares) but for finding confidence intervals for parameters estimates or for a model validation a concrete form of errors distribution is assumed. Details on maximum likelihood, least squares and related methods, such as prediction error method and instrumental variables, can be found in [3], 1] 11] 16] On the other hand, assumption concerning a functional form of error distribution are rather rarely verified. Even if the verification is done, it is usually based on residuals, i.e. differences between actual measurements and output produced by the estimated model. Such approach ....

Payne L. Goodwin, C.C. Dynamic System Identification Experiment Design and Data Analysis. Academic Press, London, 1977.

....is optimal design of inputs (or experiments ) for model identification. The problem of optimally planning experiments for inferring unknown parameters has been extensively treated in statistics, see e.g. Fedorow [1972] and in the automatic control and system theoretic literature [Mehra, 1974; Goodwin and Payne, 1977; see also Ljung, 1987] Optimal design of sensors for identification has been particularly well studied in the context of distributed parameter systems (see e.g. the survey paper of Kubrusly and Malebranche [1985] The optimal sensor location problem has been studied for detecting sensor or ....

Goodwin, G.C., and Payne, R.L.: "Dynamic System Identification: Experiment Design and Data Analysis", Academic Press, New York, 1977.

....signal design is to distribute a given amount of signal power among the harmonic components so that the experiment is optimal in some sense. Usually, the quality of the identified model is characterized by a scalar function of the Fisher information matrix F of the estimated parameters P [1, 2]. The commonly used method for optimal excitation signal design is based on the so called dispersion function [1, 2, 3] Semidefinite programming or more generally, convex optimization is well suited to the above optimal excitation signal design, since both the constraint set and the ....

....is optimal in some sense. Usually, the quality of the identified model is characterized by a scalar function of the Fisher information matrix F of the estimated parameters P [1, 2] The commonly used method for optimal excitation signal design is based on the so called dispersion function [1, 2, 3]. Semidefinite programming or more generally, convex optimization is well suited to the above optimal excitation signal design, since both the constraint set and the typical functions to be optimized are convex. 2. OPTIMAL EXCITATION SIGNAL DESIGN We apply a multi sine excitation signal ....

[Article contains additional citation context not shown here]

Goodwin, G. and R. L. Payne, Dynamic System Identification: Experiment Design and Data Analysis. Academic Press, New York, 1977.

.... is input signal dependent) Actually, this latter term brings high contributions at the time instants for which the input signal excites the i th eigen subspace determined via the corresponding eigenvector V (i) This is in contrast with classical excitations (i.e. pseudo random input signals [3, 4]) for which s 2 i can be expressed as Ns 2 0i with s 2 0i constant, for the Phi (i) V s are stationary over the data set length N . 4.2 Eigen parameters : V (N) Assuming that e(t) is a Gaussian white noise (i.e. N(0; oe 2 ) and that the excitation assumption (16) holds, we can ....

G.C. Goodwin and R.L. Payne, Dynamic System Identification : Experiment Design and Data Analysis, Mathematics in Science and Engineering, Academic Press, Vol. 136, 1977, pp 124-207.

....closed loop) was another important topic in this period. Both could well be treated within the statistical framework: simply put, it is a question of computing and analysing the Fisher Information matrix. An influential textbook from this period, with a particular emphasis on experiment design, is (Goodwin and Payne 1977). 6.4 1985 : Emerging new ideas without statistical roots. At the mid 1980:ies the statistical view of System Identification had matured and settled. The traditional and classical framework of parameter estimation had been succesfully and coherently ported to the world of dynamic systems. I ....

Goodwin, G. C. and R. L. Payne (1977). Dynamic System Identification: Experiment Design and Data Analysis. Academic Press. New York.

No context found.

G. C. Goodwin and R. L. Payne. Dynamic system identification: experiment design and data analysis. Academic Press, 1977.

No context found.

G.C. Goodwin and R.L. Payne, Dynamic System Identification: Experiment Design and Data Analysis, Academic Press, New York, 1977.

No context found.

G. C. Goodwin and L. P. Robert. Dynamic System Identification: Experiment Design and Data Analysis. Academic Press, Inc., 1977.

No context found.

G. C. Goodwin and R. L. Payne, Dynamic System Identification: Experiment Design and Data Analysis, New York, Academic Press, 1977.

No context found.

G. C. Goodwin and R. L. Payne. Dynamic system identification: experiment design and data analysis. Academic Press, 1977.

No context found.

Goodwin, G.G., and Payne, L., 1977, Dynamic System Identification Experiment Design and Data Analysis (London: Academic Press).

No context found.

Goodwin, G.C., Payne, R.L., " Dynamic System Identification : Experiment Design and Data Analysis ", Academic Press, 1977

Online articles have much greater impact   More about CiteSeer.IST   Add search form to your site   Submit documents   Feedback  

CiteSeer.IST - Copyright Penn State and NEC