Showing papers on "System identification published in 2003"

System Identification: A Frequency Domain Approach

Abstract: Preface to the First Edition Preface to the Second Edition Acknowledgments List of Operators and Notational Conventions List of Symbols List of Abbreviations Chapter 1 An Introduction to Identification Chapter 2 Measurement of Frequency Response Functions Standard Solutions Chapter 3 Frequency Response Function Measurements in the Presence of Nonlinear Distortions Chapter 4 Detection, Quantification, and Qualification of Nonlinear Distortions in FRF Measurements Chapter 5 Design of Excitation Signals Chapter 6 Models of Linear Time-Invariant Systems Chapter 7 Measurement of Frequency Response Functions The Local Polynomial Approach Chapter 8 An Intuitive Introduction to Frequency Domain Identification Chapter 9 Estimation with Know Noise Model Chapter 10 Estimation with Unknown Noise Model Standard Solutions Chapter 11 Model Selection and Validation Chapter 12 Estimation with Unknown Noise Model The Local Polynomial Approach Chapter 13 Basic Choices in System Identification Chapter 14 Guidelines for the User Chapter 15 Some Linear Algebra Fundamentals Chapter 16 Some Probability and Stochastic Convergence Fundamentals Chapter 17 Properties of Least Squares Estimators with Deterministic Weighting Chapter 18 Properties of Least Squares Estimators with Stochastic Weighting Chapter 19 Identification of Semilinear Models Chapter 20 Identification of Invariants of (Over) Parameterized Models References Subject Index Author Index About the Authors

Dynamic Textures

Abstract: Dynamic textures are sequences of images of moving scenes that exhibit certain stationarity properties in times these include sea-waves, smoke, foliage, whirlwind etc. We present a characterization of dynamic textures that poses the problems of modeling, learning, recognizing and synthesizing dynamic textures on a firm analytical footing. We borrow tools from system identification to capture the “essence” of dynamic texturess we do so by learning (i.e. identifying) models that are optimal in the sense of maximum likelihood or minimum prediction error variance. For the special case of second-order stationary processes, we identify the model sub-optimally in closed-form. Once learned, a model has predictive power and can be used for extrapolating synthetic sequences to infinite length with negligible computational cost. We present experimental evidence that, within our framework, even low-dimensional models can capture very complex visual phenomena.

Estimation with lossy measurements: jump estimators for jump systems

Abstract: In this paper, we consider estimation with lossy measurements. This problem can arise when measurements are communicated over wireless channels. We model the plant/measurement loss process as a Markovian jump linear system. While the time-varying Kalman estimator (TVKE) is known to be optimal, we introduce a simpler design, termed a jump linear estimator (JLE), to cope with losses. A JLE has predictor/corrector form, but at each time selects a corrector gain from a finite set of precalculated gains. The motivation for the JLE is twofold. The computational burden of the JLE is less than that of the TVKE and the estimation errors expected when using JLE provide an upper bound for those expected when using TVKE. We then introduce a special class of JLE, termed finite loss history estimators (FLHE), which uses a canonical gain selection logic. A notion of optimality for the FLHE is defined and an optimal synthesis method is given. The proposed design method is compared to TVKE in a simulation study.

Continuous-time model identification from sampled data: Implementation issues and performance evaluation

Abstract: This paper deals with equation error methods that fit continuous-time transfer function models to discrete-time data recently included in the CONTSID (CONtinuous-Time System IDentification) Matlab toolbox. An overview of the methods is first given where implementation issues are highlighted. The performances of the methods are then evaluated on simulated examples by Monte Carlo simulations. The experiments have been carried out to study the sensitivity of each approach to the design parameters, sampling period, signal-to-noise ratio, noise power spectral density and type of input signal. The effectiveness of the CONTSID toolbox techniques is also briefly compared with indirect methods in which discrete-time models are first estimated and then transformed into continuous-time models. The paper does not consider iterative or recursive algorithms for continuous-time transfer function model identification.

System identification using binary sensors

Abstract: System identification is investigated for plants that are equipped with only binary-valued sensors. Optimal identification errors, time complexity, optimal input design, and impact of disturbances and unmodeled dynamics on identification accuracy and complexity are examined in both stochastic and deterministic information frameworks. It is revealed that binary sensors impose fundamental limitations on identification accuracy and time complexity, and carry distinct features beyond identification with regular sensors. Comparisons between the stochastic and deterministic frameworks indicate a complementary nature in their utility in binary-sensor identification.

A class of frequency-domain adaptive approaches to blind multichannel identification

Abstract: We extend our previous studies on adaptive blind channel identification from the time domain into the frequency domain. A class of frequency-domain adaptive approaches, including the multichannel frequency-domain LMS (MCFLMS) and constrained/unconstrained normalized multichannel frequency-domain LMS (NMCFLMS) algorithms, are proposed. By utilizing the fast Fourier transform (FFT) and overlap-save techniques, the convolution and correlation operations that are computationally intensive when performed by the time-domain multichannel LMS (MCLMS) or multichannel Newton (MCN) methods are efficiently implemented in the frequency domain, and the MCFLMS is rigorously derived. In order to achieve independent and uniform convergence for each filter coefficient and, therefore, accelerate the overall convergence, the coefficient updates are properly normalized at each iteration, and the NMCFLMS algorithms are developed. Simulations show that the frequency-domain adaptive approaches perform as well as or better than their time-domain counterparts and the cross-relation (CR) batch method in most practical cases. It is remarkable that for a three-channel acoustic system with long impulse responses (256 taps in each channel) excited by a male speech signal, only the proposed NMCFLMS algorithm succeeds in determining a reasonably accurate channel estimate, which is good enough for applications such as time delay estimation.

Identification of nonlinear physiological systems

Abstract: Preface. 1. Introduction. 1.1 Signals. 1.2 Systems and Models. 1.3 System Modeling. 1.4 System Identification. 1.5 How Common are Nonlinear Systems? 2. Background. 2.1 Vectors and Matrices. 2.2 Gaussian Random Variables. 2.3 Correlation Functions. 2.4 Mean-Square Parameter Estimation. 2.5 Polynomials. 2.6 Notes and References. 2.7 Problems. 2.8 Computer Exercises. 3. Models of Linear Systems. 3.1 Linear Systems. 3.2 Nonparametric Models. 3.3 Parametric Models. 3.4 State-Space Models. 3.5 Notes and References. 3.6 Theoretical Problems. 3.7 Computer Exercises. 4. Models of Nonlinear Systems. 4.1 The Volterra Series. 4.2 The Wiener Series. 4.3 Simple Block Structures. 4.4 Parallel Cascades. 4.5 The Wiener-Bose Model. 4.6 Notes and References. 4.7 Theoretical Problems. 4.8 Computer Exercises. 5. Identification of Linear Systems. 5.1 Introduction. 5.2 Nonparametric Time-Domain Models. 5.3 Frequency Response Estimation. 5.4 Parametric Methods. 5.5 Notes and References. 5.6 Computer Exercises. 6. Correlation-Based Methods. 6.1 Methods for Functional Expansions. 6.2 Block Structured Models. 6.3 Problems. 6.4 Computer Exercises. 7. Explicit Least-Squares Methods. 7.1 Introduction. 7.2 The Orthogonal Algorithms. 7.3 Expansion Bases. 7.4 Principal Dynamic Modes. 7.5 Problems. 7.6 Computer Exercises. 8. Iterative Least-Squares Methods. 8.1 Optimization Methods. 8.2 Parallel Cascade Methods. 8.3 Application: Visual Processing in the Light Adapted Fly Retina. 8.4 Problems 8.5 Computer Exercises. References. Index. IEEE Press Series in Biomedical Engineering.

Recursive identification of Hammerstein systems with discontinuous nonlinearities containing dead-zones

Abstract: This note deals with the recursive parameter identification of Hammerstein systems with discontinuous nonlinearities, i.e., two-segment piecewise-linear with dead-zones and preloads. A special form of the Hammerstein model with this type of nonlinearity is incorporated into the recursive least squares identification scheme supplemented with the estimation of model internal variables. The proposed method is illustrated by examples.

Bayesian Fast Fourier Transform Approach for Modal Updating Using Ambient Data

Abstract: The problem of identification of the modal parameters of a structural model using measured ambient response time histories is addressed. A Bayesian Fast Fourier Transform approach (BFFTA) for modal updating is presented which uses the statistical properties of the Fast Fourier transform (FFT) to obtain not only the optimal values of the updated modal parameters but also their associated uncertainties, calculated from their joint probability distribution. Calculation of the uncertainties of the identified modal parameters is very important when one plans to proceed with the updating of a theoretical finite element model based on modal estimates. The proposed approach requires only one set of response data in contrast to many of the existing frequency-based approaches which require averaging. It is found that the updated PDF can be well approximated by a Gaussian distribution centred at the optimal parameters at which the posterior PDF is maximized. Examples using simulated data are presented to illustrate ...

Simulation Methods for Optimal Experimental Design in Systems Biology

Abstract: To obtain a systems-level understanding of a biological system, the authors conducted quantitative dynamic experiments from which the system structure and the parameters have to be deduced. Since biological systems have to cope with different environmental conditions, certain properties are often robust with respect to variations in some of the parameters. Hence, it is important to use optimal experimental design considerations in advance of the experiments to improve the information content of the measurements. Using the MAP-Kinase pathway as an example, the authors present a simulation study investigating the application of different optimality criteria. It is demonstrated that experimental design significantly improves the parameter estimation accuracy and also reveals difficulties in parameter estimation due to robustness.

Mistuning Identification of Bladed Disks Using a Fundamental Mistuning Model—Part I: Theory

Abstract: This paper is the first in a two-part study of identifying mistuning in bladed disks. It develops a new method of mistuning identification based on measurements of the vibratory response of the system as a whole. As a system-based method, this approach is particularly suited to integrally bladed rotors, whose blades cannot be removed for individual measurements. The method is based on a recently developed reduced order model of mistuning called the Fundamental Mistuning Model, FMM, and is applicable to isolated families of modes. Two versions of FMM system identification are presented: a basic version that requires some prior knowledge of the system’s properties, and a somewhat more complex version that determines the mistuning completely from experimental data.Copyright © 2003 by ASME

Frequency domain identification of Hammerstein models

Abstract: Discusses Hammerstein model identification in the frequency domain using sampled input-output data. By exploring the fundamental frequency and harmonics generated by the unknown nonlinearity, we propose a frequency domain approach and show its convergence for both the linear and nonlinear subsystems in the presence of noise. No a priori knowledge of the structure of the nonlinearity is required and the linear part can be nonparametric.

Neural Network Control

Abstract: This thesis addresses two neural network based control systems. The first is a neural network based predictive controller. System identification and controller design are discussed. The second is a direct neural network controller. Parameter choice and training methods are discussed. Both controllers are tested on two different plants. Problems regarding implementations are discussed. First the neural network based predictive controller is introduced as an extension to the generalised predictive controller (GPC) to allow control of non-linear plant. The controller design includes the GPC parameters, but prediction is done explicitly by using a neural network model of the plant. System identification is discussed. Two control systems are constructed for two different plants: A coupled tank system and an inverse pendulum. This shows how implementation aspects such as plant excitation during system identification are handled. Limitations of the controller type are discussed and shown on the two implementations. In the second part of this thesis, the direct neural network controller is discussed. An output feedback controller is constructed around a neural network. Controller parameters are determined using system simulations. The control system is applied as a single-step ahead controller to two different plants. One of them is a path-following problem in connection with a reversing trailer truck. This system illustrates an approach with step-wise increasing controller complexity to handle the unstable control object. The second plant is a coupled tank system. Comparison is made with the first controller. Both controllers are shown to work. But for the neural network based predictive controller, construction of a neural network model of high accuracy is critical especially when long prediction horizons are needed. This limits application to plants that can be modelled to sufficient accuracy. The direct neural network controller does not need a model. Instead the controller is trained on simulation runs of the plant. This requires careful selection of training scenarios, as these scenarios have impact on the performance of the controller.

Subspace identification with guaranteed stability using constrained optimization

Abstract: In system identification, the true system is often known to be stable. However, due to finite sample constraints, modeling errors, plant disturbances and measurement noise, the identified model may be unstable. We present a constrained optimization method to ensure asymptotic stability of the identified model in the context of subspace identification methods. In subspace identification, we first obtain an estimate of the state sequence or extended observability matrix and then solve a least squares optimization problem to estimate the system parameters. To ensure asymptotic stability of the identified model, we write the least-squares optimization problem as a convex linear programming problem with mixed equality, quadratic, and positive-semidefinite constraints suitable for existing convex optimization codes such as SeDuMi. We present examples to illustrate the method and compare to existing approaches.

Local and Piecewise Affine Approaches to System Identification

Abstract: Identification of nonlinear systems is a multifaceted research area, with many diverse approaches and methods. This thesis considers two different approaches: (nonparametric) local modelling, and i ...

Reconfigurable Flight Control System Design Using Direct Adaptive Method

Abstract: A reconfigurable flight control system provides better survivability through the automatic reconfiguration of control system when faults occur during flight. The adaptive control method has been effectively applied to the reconfigurable flight control system design. However, reconfigurable flight control systems based on the indirect adaptive control method require persistent input excitation and smooth input-output data. To deal with the persistent input excitation problem and to obtain smooth control input, the system identification algorithm in reconfiguration flight control systems usually imposes some constraints on past input-output data, which may deteriorate the reconfiguration performance. Thus, a new reconfigurable flight control system based on the direct adaptive method is proposed to achieve better reconfiguration performance without the system identification process. The proposed control method uses a model following controller with direct adaptive update rules. To control the inner-loop states and the outer-loop states of the flight system simultaneously,the timescale separation principle is applied. The reconfiguration performance of the proposed control method is evaluated through numerical simulations using a six-degree-of-freedom nonlinear aircraft model.

Spatial Control of Vibration: Theory and Experiments

Abstract: Modelling spatial norms and model reduction model correction spatial control optimal placement of actuators and sensors system identification for spatially distributed systems.