Showing papers on "System identification published in 1980"
TL;DR: In this paper, an unknown linear time-invariant system without control, driven by a white noise with known distribution, is considered, and the identification of both gain and phase of the system, observing only the output, is presented.
Abstract: Consider an unknown linear time-invariant system without control, driven by a white noise with known distribution. We are interested in the identification of this system, observing only the output. This problem is well known under the major assumption: the system is minimum (or maximum!) phase, in which the very popular least squares method gives an identification of the system in an autoregressive form. However, we are Interested in the case where the system is nonminimum (nor maximum!) phase, i.e., we want identification of both gain and phase of the system. The literature gives only a negative result: the idenfication of the phase of the system is impossible in the case of a Gaussian driving noise (hence, second-order statistics are irrelevant to our problem). For a large class of other input distributions, we present an identification procedure, and give some numerical results for a concrete case origin of our study: the blind adjustment of a transversal equalizer without any startup period prior to data transmission.
517 citations
TL;DR: The basic ideas behind the parameter estimation methods are discussed in a general setting and an example is given which illustrates some properties of the methods and shows the usefulness of interactive computing.
372 citations
01 Jan 1980
TL;DR: In this article, two methods of system identification of n degrees-of-freedom structural dynamic systems are studied and applied to identification of the hydrodynamic coefficient matrices associated with nonlinear drag and linear inertia forces appearing in the equations of motion of offshore structures subjected to wave forces.
Abstract: Two methods of system identification of n degrees-of-freedom nonlinear structural dynamic systems are studied and applied to identification of the hydrodynamic coefficient matrices associated with nonlinear drag and linear inertia forces appearing in the equations of motion of offshore structures subjected to wave forces. These two methods, being essentially the methods of state estimation, use nonlinear Kalman filtering algorithms which can be applied to parameter estimation problems by regarding each of the parameters involved in the system as an augmented state variable. One of these methods uses the extended Kalman filter while the other was the iterated linear filter-smoother. Analytical simulation studies are performed for two degrees-of-freedom structural systems on the basis of artificially generated input and output observations under the various output noise conditions. Both methods yield good estimates even under the conditions of fairly large amounts of output noise and are moderately...
277 citations
TL;DR: In this paper, the problem of determining linear models of structures from seismic response data is investigated using ideas from the theory of system identification, where the approach is to determine the optimal estimates of the model parameters by minimizing a selected measure-of-fit between the responses of the structure and the model.
Abstract: The problem of determining linear models of structures from seismic response data is investigated using ideas from the theory of system identification. The approach is to determine the optimal estimates of the model parameters by minimizing a selected measure-of-fit between the responses of the structure and the model. Because earthquake records are normally available from only a small number of locations in a structure, and because of noise in the records, it is necessary in practice to estimate parameters of the dominant modes in the records, rather than the stiffness and damping matrices of the linear model. A new algorithm is developed to determine the optimal estimates of the modal parameters. After tests with simulated data, the method is applied to a multi-storey building using records from the 1971 San Fernando earthquake in California. New information is obtained concerning the properties of the lower modes of the building and the time-varying character of the equivalent linear parameters.
190 citations
01 Jan 1980
187 citations
01 Jan 1980
TL;DR: In this paper, a general class of parameter estimation methods for stochastic dynamical systems is studied and the class contains the least squares method, output-error methods, the maximum likelihood method and several other techniques.
Abstract: A general class of parameter estimation methods for stochastic dynamical systems is studied. The class contains the least squares method, output-error methods, the maximum likelihood method and several other techniques. It is shown that the class of estimates so obtained are asymptotically normal and expressions for the resulting asymptotic covariance matrices are given. The regularity conditions that are imposed to obtain these results, are fairly weak. It is, for example, not assumed that the true system can be described within the chosen model set, and, as a consequence, the results in this paper form a part of the so-called approximate modeling approach to system identification. It is also noteworthy that arbitrary feedback from observed system outputs to observed system inputs is allowed and stationarity is not required
162 citations
TL;DR: In this tutorial very simple methods will be presented by which models for dynamic processes can be obtained that employ step responses of the process or responses to other nonperiodic testsignals or process responses to periodic signals.
141 citations
TL;DR: This article demonstrates the application of least squares for the estimation of system parameters and solutions are discussed for the case of white noise and correlated noise corrupting the useful output signal of the system.
123 citations
TL;DR: A general procedure of process identification, the selection of input signals, theselection of the sampling time, off-line and on-line identification, comparison of parameter estimation methods, model order testing and model verification is presented.
106 citations
TL;DR: In this article, the validity of the AIC approach to a class of time-varying AR models is investigated. But the AKAKE criterion has been applied extensively to constant coefficient AR models and the question of validity of AIC for more general models appears to be open.
Abstract: We are often faced with the problem of estimating the number of parameters, as well as their values, for a model that will fit a given set of observations. An important study of this identification problem was made by Akaike, who improved the maximum likelihood principle and established the so-called AIC criterion to select suitable models. This criterion has been applied extensively to constant coefficient AR models. The question of validity of the AIC approach for more general models appears to be open. Motivated by the desire to model nonstationary geophysical data the main purpose of this paper is to establish conditions guaranteeing the validity of the AIC approach to a class of time-varying AR models.
66 citations
TL;DR: In this paper, the effects of noise in the data are neglected in deriving these equations, which can lead to very large errors in the estimates, and a new set of transforms based on the idea of maximizing their Fisher information content is developed.
Abstract: Linear transform methods like moments, modulating functions, and Laplace transforms are widely used for parameter estimation in system identification problems because they can reduce a large set of overdetermined equations to a small set of linear and nonlinear equations, which often have a very simple form and a unique solution. However, the effects of noise in the data are neglected in deriving these equations. We show (in terms of Fisher's information measure, the generalized variance, and simulations) that these methods can lead to very large errors in the estimates. We develop a new set of transforms based on the idea of maximizing their Fisher information content. The robustness of these new transforms, in contrast to the others, is illustrated by simulations of nanosecond flourescence decay and multicomponent exponential decay.
TL;DR: The application of optimal experimental design theory to models for dynamic systems is surveyed and applications are split roughly into those involving choice of input functions and those in which sampling rates also are selected.
Abstract: The application of optimal experimental design theory to models for dynamic systems is surveyed. Preliminary sections briefly discuss the models used and the main points of statistical optimal design theory. Then the ways in which the latter carry over to dynamic models are described. These applications are split roughly into those involving choice of input functions and those in which sampling rates also are selected.
01 Jan 1980
TL;DR: This paper surveys sequential filter adaptation techniques and some applications for transversal FIR, lattice and recursive filters, which span a wide spectrum of possible performance/complexity tradeoffs.
Abstract: Over the past few years a number of new adaptive filter algorithms have been developed and applied to meet demands for faster convergence and better tracking properties than earlier techniques could offer Applications include adaptive channel equalization, adaptive predictive speech coding and on-line system identification This paper surveys sequential filter adaptation techniques and some applications for transversal FIR, lattice and recursive filters The available techniques fit into two main categories: (1) gradient-type methods (exemplified by the well-known LMS algorithm) in which successive corrections to adaptive system parameters are only correct in an average sense, and (2) recursive least-squares methods, which continuously provide the solution to a numerical optimization problem, given all the preceding data The available techniques span a wide spectrum of possible performance/complexity tradeoffs
Patent•
20 Oct 1980
TL;DR: In this paper, a P-I-D controller in a process control system is self tuned in response to discrete time model identification parameters which are converted to P-II-D form.
Abstract: A P-I-D controller in a process control system is self tuned in response to discrete time model identification parameters which are converted to P-I-D form. Additionally, the P-I-D parameters are limited to prevent them from exceeding the controller's specified range.
TL;DR: This paper discusses several important aspects of the implementation of a short-time spectral analysis approach to the problems of spectral estimation and system identification.
Abstract: Recent work has demonstrated the utility of a short-time spectral analysis approach to the problems of spectral estimation and system identification. In this paper several important aspects of the implementation are discussed. Included is a discussion of the computational effects (e.g., storage, running time) of the various analysis parameters. A computer program is included which illustrates one implementation of the method.
TL;DR: This paper presents a basic introduction to model building and identification, both fundamentals of mathematical description of dynamic systems and the principles of theoretical modelling are emphasized.
TL;DR: In this paper, the problem of finding a (minimal) compensator S 3 such that the transfer matrices of S 1 and S 2 have the same first N Markov matrices is considered and solved.
Abstract: Given two systems S 1 and S 2 , the problem of finding a (minimal) compensator S 3 such that the transfer matrices of S_{1}S_{3} and S 2 have the same first N Markov matrices is considered and solved.
TL;DR: In this article, a pencil-of-functions system identification technique for weakly nonlinear systems is presented. But the identification technique is a black-box procedure in that only measurements at the system input and output terminals are required.
Abstract: A new identification technique for a class of weakly nonlinear systems whose behavior is adequately characterized in terms of a finite Volterra functional series is presented. Application of the identification technique results in a complete specification of the nonlinear impulse responses which describe a weakly nonlinear system. The identification technique is a "black box" procedure in that only measurements at the system input and output terminals are required. A functional form for the second-order impulse response, h_2(t_1, t_2) , is derived for a nonlinear system with a finite number of power-law devices. The identification of h_2(t_1, t_2) is accomplished by exciting the system with a sum of exponentially decaying signals and appropriately processing the input and output signals using the pencil-of-functions system identification approach. This results in a complete set of linear equations involving all the parameters of h_2(t_1, t_2) . Solution of these equations uniquely determines h_2(t_l, t_2) . An example of the practical application of the technique to a common emitter amplifier is presented.
TL;DR: The role played by scale models in dynamical systems studies and control engineering is described in this article, where the development at the Control Systems Centre, UMIST, of a series of scale models for control systems studies is described.
Abstract: The article describes the role played by scale models in dynamical systems studies and control engineering. As a tangible example, the development at the Control Systems Centre, UMIST, of a series of scale models for control systems studies is described. The models form a comprehensive set of equipment which is used for laboratory investigations of process systems, material transport problems, electromechanical control, and the like. The paper describes how such a range of models is used to illustrate the essential features and methods of mathematical modelling, system identification, analysis and control
TL;DR: In this article, a wide variety of parameter estimation techniques can be discussed from the point of view of functional operators working on system input/output signals, which can be characterised by time functions, called ''template functions?''.
Abstract: A wide variety of parameter estimation techniques can be discussed from the point of view of functional operators working on system input/output signals. The classes of operators can be characterised by time functions, called `template functions?. These notions contribute to a coherent picture of a wide class of estimation techniques, as well as to a discussion of the statistical properties and, in appropriate cases, the information efficiency.
TL;DR: In this article, it is shown that the equations of motion of an aeroelastic system may be derived from measured response data, which can then be used to calculate dynamic characteristics of the system at any chosen values of kinetic pressure.
01 Jan 1980
TL;DR: The Response Surface Methodology (RSM) as mentioned in this paper is a general approach to system identification that organises several statistical techniques in order to provide an estimate of the p.d.
Abstract: The Response Surface Methodology as a general approach to the system identification is presented. The method organises several statistical techniques in order to provide an estimate of the p.d.f. of the output variable of the identified system as a function of the p.d.f.’s of the input variable, as the final result. In particular the following techniques are dealt with:
the sensitivity analysis;
the choice of the approximating function;
the experimental design;
the parameter estimation.
A typical application is provided.
TL;DR: In this article, the family of m -input, n-dimensional linear systems can be globally identified with a generic input sequence of length 2mn, which is the best possible.
Abstract: The family of m -input, n -dimensional linear systems can be globally identified with a generic input sequence of length 2mn . This bound is the best possible. A best bound is provided also for a corresponding local identification problem.
TL;DR: Akaike's method for model identification has been used to identify Markov chain models for simple transformations of daily precipitation at three locations in southern Norway, and wind force and wave height at one location in the Norwegian Sea.
Abstract: Akaike's method for model identification has been used to identify Markov chain models for simple transformations of daily precipitation at three locations in southern Norway, and wind force and wave height at one location in the Norwegian Sea. Attempts at identification of the horizontal wind vector as an autoregressive process also have been made. The estimated order of a model appears to increase with the sample size. It also may have a significant uncertainty. The analytical complexity of identified models may appear to be unnecessarily large for some purposes.
TL;DR: In this article, the sensitivity of parameter estimates to observation noise or disturbances affects the accuracy of the parameter estimates in system identification, and it is shown that the (parameter estimate)/observation sensitivity tends to be lowered for optimal inputs.
TL;DR: This paper considers the problem of identifying multiinput single-output linear time-invariant discrete-time systems from noise-free input/output measurements and shows that the correctness associated with an identification is critically dependent on the input sequence used.
Abstract: This paper considers the problem of identifying multiinput single-output linear time-invariant discrete-time systems from noise-free input/output measurements The effect or input on identification is studied in detail It is shown that the correctness associated with an identification is critically dependent on the input sequence used Furthermore, as long as only finite input/output sequences are used for identification, there is always a possibility that the actual system is of higher order than the identified models reveal This uncertainty leads to the concept of tentative identification Sufficient conditions for correct order determination and modeling are also investigated
TL;DR: In this article, the propriety of adopting a multi-degree-of-freedom lumped mass-spring-dampers system driven by white noise support excitation as a one-dimensional model for a soil-layer-bedrock system during an earthquake is investigated by means of statistical system identification of the model with noisy measurement of the earthquake ground velocity.
Abstract: The propriety of adopting a multi-degree-of-freedom lumped mass–spring–dampers system driven by white noise support excitation as a one-dimensional model for a soil-layer–bedrock system during an earthquake is investigated by means of statistical system identification of the model with noisy measurement of the earthquake ground velocity. The present discussion also suggests that this model may not be applicable to all observed earthquake records, since the model itself depends on the statistical nature of the earthquake motion. For appropriate earthquake records, the system identification procedure may be accomplished; then dynamical properties of the soil-layer and the power spectral density for white noise excitation acting upon the bedrock can be estimated as shown in a numerical example.
01 Dec 1980
TL;DR: In this paper, a method for automatically controlling the voltage of a Gas-Tungsten-Arc-Welding (GTAW) process utilize an analog servo system to regulate the welding electrode to workpiece distance.
Abstract: Present methods for automatically controlling the voltage of a Gas-Tungsten-Arc-Welding (GTAW) process utilize an analog servo system to regulate the welding electrode to workpiece distance. The controllers currently being used in this facility are characterized by large over-shoot to a step input, and steady-state error due to deadband. An attempt to use microprocessor technology to improve the performance is underway. Such a method of improving the performance of the voltage controllers has been developed using an on-line identification procedure to determine the mathematical model of the present controller. Once an adequate model has been determined, a digital compensation scheme is designed to acquire the desired performance. The identification procedure used was the Model Reference Adaptive Technique, by which a model is postulated and its performance is compared to the actual plant performance for a set of given inputs. The squared error between the plant performance and the predicted performance is then computed over the identification interval. The model parameters are adjusted by an optimization routine to improve the model of the plant. Once the model is optimized to the point of diminishing returns, a digital PI or PID controller is designed. The result is that now the machine behaves in a manner that can be easily adjusted to give significantly improved performance. A relatively inexpensive M6800 microcomputer is used for both the on-line model identification, including the error minimizing routines, and for the realization of the series compensating controller. The software used is Motorola 6800 Assembly Language and Motorola Fortran. The method of improved control is being expanded to include the other welding parameters (current and speed of torch travel) and to other methods of control and state reconstruction, such as the linear regulator and the Kalman filter. Performance comparisons between the linear regulator and PID controller are made.
TL;DR: The emphasis is on the design, or more correctly, the synthesis, of the controller, and classes of optimal controllers each dependent on a different cost function are discussed.
Abstract: Self- tuning regulators and controllers have proven to be a useful industrial control device. The self- tuning theory upon which these devices are based comprises the two aspects: controller design and system identification. In this first paper the emphasis is on the design, or more correctly, the synthesis, of the controller. In particular, classes of optimal controllers each dependent on a different cost function are discussed. The two most popular controllers- the minimum variance and the generalised minimum variance control laws - are derived in this first part.