scispace - formally typeset
Search or ask a question
Topic

System identification

About: System identification is a research topic. Over the lifetime, 21291 publications have been published within this topic receiving 439142 citations.


Papers
More filters
Journal ArticleDOI
TL;DR: The aim of this work is to provide new insights on the stable spline estimator equipped with ML estimation of hyperparameters, and to derive the notion of excess degrees of freedom, which measures the additional complexity to be assigned to an estimator which is also required to determinehyperparameters from data.

93 citations

Journal ArticleDOI
TL;DR: In this paper, a structural damage detection method based on parameter identification using an iterative neural network (NN) technique is proposed, which is first trained off-line using an initial training data set that consists of assumed structural parameters as outputs and their corresponding dynamic characteristics as inputs.
Abstract: A structural damage detection method based on parameter identification using an iterative neural network (NN) technique is proposed in this study. The NN model is first trained off-line using an initial training data set that consists of assumed structural parameters as outputs and their corresponding dynamic characteristics as inputs. The structural parameters are assumed with different levels of reduction to simulate various degrees of structural damage. The concept of orthogonal array is adopted to generate the representative combinations of parameter changes, which can significantly reduce the number of training data while maintaining the data completeness. A modified back-propagation learning algorithm is proposed which can overcome possible saturation of the sigmoid function and speed up the training process. The trained NN model is used to predict the structural parameters by feeding in measured dynamic characteristics. The predicted structural parameters are then used in the FE model to calculate ...

92 citations

Journal ArticleDOI
TL;DR: In this paper, a relaxed orthonormal vector fitting (ROVF) is proposed for frequency-domain responses in electromagnetic transients programs for frequencydependent modeling of transmission lines and to some extent, FDNEs and transformers.
Abstract: Rational approximation of frequency-domain responses is commonly used in electromagnetic transients programs for frequency-dependent modeling of transmission lines and to some extent, network equivalents (FDNEs) and transformers. This paper analyses one of the techniques [vector fitting (VF)] within a general iterative least-squares scheme that also explains the relation with the polynomial-based Sanathanan-Koerner iteration. Two recent enhancements of the original VF formulation are described: orthonormal vector fitting (OVF) which uses orthonormal functions as basis functions instead of partial fractions, and relaxed vector fitting (RVF), which uses a relaxed least-squares normalization for the pole identification step. These approaches have been combined into a single approach: relaxed orthonormal vector fitting (ROVF). The application to FDNE identification shows that ROVF offers more robustness and better convergence than the original VF formulation. Alternative formulations using explicit weighting and total least squares are also explored.

92 citations

Journal ArticleDOI
TL;DR: In this paper, a system identification methodology is presented for determining the values of control parameters in a continuously smooth hysteretic model for inelastic dynamic behavior of structural concrete systems.

92 citations

Dissertation
01 Jan 1997
TL;DR: The work described in this thesis addresses this fundamental problem with neurofuzzy modelling, as a result algorithms which are less sensitive to the quality of the a priori knowledge and empirical data are developed.
Abstract: System identification is the task of constructing representative models of processes and has become an invaluable tool in many different areas of science and engineering. Due to the inherent complexity of many real world systems the application of traditional techniques is limited. In such instances more sophisticated (so called intelligent) modelling approaches are required. Neurofuzzy modelling is one such technique, which by integrating the attributes of fuzzy systems and neural networks is ideally suited to system identification. This attractive paradigm combines the well established learning techniques of a particular form of neural network i.e. generalised linear models with the transparent knowledge representation of fuzzy systems, thus producing models which possess the ability to learn from real world observations and whose behaviour can be described naturally as a series of linguistic humanly understandable rules. Unfortunately, the application of these systems is limited to low dimensional problems for which good quality expert knowledge and data are available. The work described in this thesis addresses this fundamental problem with neurofuzzy modelling, as a result algorithms which are less sensitive to the quality of the a priori knowledge and empirical data are developed. The true modelling capabilities of any strategy is heavily reliant on the model's structure, and hence an important (arguably the most important) task is structure identification. Also, due to the curse of dimensionality, in high dimensional problems the size of conventional neurofuzzy models gets prohibitively large. These issues are tackled by the development of automatic neurofuzzy model identification algorithms, which exploit the available expert knowledge and empirical data. To alleviate problems associated with the curse of dimensionality, aid model generalisation and enhance model transparency, parsimonious models are identified. This is achieved by the application of additive and multiplicative neurofuzzy models which exploit structural redundancies found in conventional systems. The developed construction algorithms successfully identify parsimonious models, but as a result of noisy and poorly distributed empirical data, these models can still generalise inadequately. This problem is addressed by the application of Bayesian inferencing techniques; a form of regularisation. Smooth model outputs are assumed and superfluous model parameters are controlled, sufficiently aiding model generalisation and transparency, and data interpolation and extrapolation. By exploiting the structural decomposition of the identified neurofuzzy models, an efficient local method of regularisation is developed. All the methods introduced in this thesis are illustrated on many different examples, including simulated time series, complex functional equations, and multi-dimensional dynamical systems. For many of these problems conventional neurofuzzy modelling is unsuitable, and the developed techniques have extended the range of problems to which neurofuzzy modelling can successfully be applied.

92 citations


Network Information
Related Topics (5)
Control theory
299.6K papers, 3.1M citations
93% related
Linear system
59.5K papers, 1.4M citations
92% related
Robustness (computer science)
94.7K papers, 1.6M citations
91% related
Control system
129K papers, 1.5M citations
91% related
Optimization problem
96.4K papers, 2.1M citations
88% related
Performance
Metrics
No. of papers in the topic in previous years
YearPapers
2023177
2022361
2021646
2020813
2019804
2018862