Showing papers on "System identification published in 2014"

Identification of Dynamic Systems: An Introduction with Applications

Abstract: Precise dynamic models of processes are required for many applications, ranging from control engineering to the natural sciences and economics. Frequently, such precise models cannot be derived using theoretical considerations alone. Therefore, they must be determined experimentally. This book treats the determination of dynamic models based on measurements taken at the process, which is known as system identification or process identification. Both offline and online methods are presented, i.e. methods that post-process the measured data as well as methods that provide models during the measurement. The book is theory-oriented and application-oriented and most methods covered have been used successfully in practical applications for many different processes. Illustrative examples in this book with real measured data range from hydraulic and electric actuators up to combustion engines. Real experimental data is also provided on the Springer webpage, allowing readers to gather their first experience with the methods presented in this book. Among others, the book covers the following subjects: determination of the non-parametric frequency response, (fast) Fourier transform, correlation analysis, parameter estimation with a focus on the method of Least Squares and modifications, identification of time-variant processes, identification in closed-loop, identification of continuous time processes, and subspace methods. Some methods for nonlinear system identification are also considered, such as the Extended Kalman filter and neural networks. The different methods are compared by using a real three-mass oscillator process, a model of a drive train. For many identification methods, hints for the practical implementation and application are provided. The book is intended to meet the needs of students and practicing engineers working in research and development, design and manufacturing.

Principles of System Identification: Theory and Practice

Abstract: Master Techniques and Successfully Build Models Using a Single Resource Vital to all data-driven or measurement-based process operations, system identification is an interface that is based on observational science, and centers on developing mathematical models from observed data. Principles of System Identification: Theory and Practice is an introductory-level book that presents the basic foundations and underlying methods relevant to system identification. The overall scope of the book focuses on system identification with an emphasis on practice, and concentrates most specifically on discrete-time linear system identification. Useful for Both Theory and Practice The book presents the foundational pillars of identification, namely, the theory of discrete-time LTI systems, the basics of signal processing, the theory of random processes, and estimation theory. It explains the core theoretical concepts of building (linear) dynamic models from experimental data, as well as the experimental and practical aspects of identification. The author offers glimpses of modern developments in this area, and provides numerical and simulation-based examples, case studies, end-of-chapter problems, and other ample references to code for illustration and training. Comprising 26 chapters, and ideal for coursework and self-study, this extensive text: Provides the essential concepts of identification Lays down the foundations of mathematical descriptions of systems, random processes, and estimation in the context of identification Discusses the theory pertaining to non-parametric and parametric models for deterministic-plus-stochastic LTI systems in detail Demonstrates the concepts and methods of identification on different case-studies Presents a gradual development of state-space identification and grey-box modeling Offers an overview of advanced topics of identification namely the linear time-varying (LTV), non-linear, and closed-loop identification Discusses a multivariable approach to identification using the iterative principal component analysis Embeds MATLAB® codes for illustrated examples in the text at the respective points Principles of System Identification: Theory and Practice presents a formal base in LTI deterministic and stochastic systems modeling and estimation theory; it is a one-stop reference for introductory to moderately advanced courses on system identification, as well as introductory courses on stochastic signal processing or time-series analysis.The MATLAB scripts and SIMULINK models used as examples and case studies in the book are also available on the author's website: http://arunkt.wix.com/homepage#!textbook/c397

System Identification Via Sparse Multiple Kernel-Based Regularization Using Sequential Convex Optimization Techniques

Abstract: Model estimation and structure detection with short data records are two issues that receive increasing interests in System Identification. In this paper, a multiple kernel-based regularization method is proposed to handle those issues. Multiple kernels are conic combinations of fixed kernels suitable for impulse response estimation, and equip the kernel-based regularization method with three features. First, multiple kernels can better capture complicated dynamics than single kernels. Second, the estimation of their weights by maximizing the marginal likelihood favors sparse optimal weights, which enables this method to tackle various structure detection problems, e.g., the sparse dynamic network identification and the segmentation of linear systems. Third, the marginal likelihood maximization problem is a difference of convex programming problem. It is thus possible to find a locally optimal solution efficiently by using a majorization minimization algorithm and an interior point method where the cost of a single interior-point iteration grows linearly in the number of fixed kernels. Monte Carlo simulations show that the locally optimal solutions lead to good performance for randomly generated starting points.

Connecting System Identification and Robust Control for Next-Generation Motion Control of a Wafer Stage

Abstract: Next-generation precision motion systems are lightweight to meet stringent requirements regarding throughput and accuracy. Such lightweight systems typically exhibit lightly damped flexible dynamics in the controller cross-over region. State-of-the-art modeling and motion control design procedures do not deliver the required model complexity and fidelity to control the flexible dynamical behavior. The aim of this paper is to develop a combined system identification and robust control design procedure for high performance motion control and apply it to a wafer stage. Hereto, new connections between system identification and robust control are employed. The experimental results confirm that the proposed procedure significantly extends existing results and enables next-generation motion control design.

System Parameter Identification: Information Criteria and Algorithms

Abstract: Recently, criterion functions based on information theoretic measures (entropy, mutual information, information divergence) have attracted attention and become an emerging area of study in signal processing and system identification domain. This book presents a systematic framework for system identification and information processing, investigating system identification from an information theory point of view. The book is divided into six chapters, which cover the information needed to understand the theory and application of system parameter identification. The authors' research provides a base for the book, but it incorporates the results from the latest international research publications. One of the first books to present system parameter identification with information theoretic criteria so readers can track the latest developmentsContains numerous illustrative examples to help the reader grasp basic methods

Online adaptive approximate optimal tracking control with simplified dual approximation structure for continuous-time unknown nonlinear systems

Abstract: This paper proposes an online adaptive approximate solution for the infinite-horizon optimal tracking control problem of continuous-time nonlinear systems with unknown dynamics. The requirement of the complete knowledge of system dynamics is avoided by employing an adaptive identifier in conjunction with a novel adaptive law, such that the estimated identifier weights converge to a small neighborhood of their ideal values. An adaptive steady-state controller is developed to maintain the desired tracking performance at the steady-state, and an adaptive optimal controller is designed to stabilize the tracking error dynamics in an optimal manner. For this purpose, a critic neural network (NN) is utilized to approximate the optimal value function of the Hamilton-Jacobi-Bellman (HJB) equation, which is used in the construction of the optimal controller. The learning of two NNs, i.e., the identifier NN and the critic NN, is continuous and simultaneous by means of a novel adaptive law design methodology based on the parameter estimation error. Stability of the whole system consisting of the identifier NN, the critic NN and the optimal tracking control is guaranteed using Lyapunov theory; convergence to a near-optimal control law is proved. Simulation results exemplify the effectiveness of the proposed method.

Camera Model Identification Based on the Heteroscedastic Noise Model

Abstract: The goal of this paper is to design a statistical test for the camera model identification problem. The approach is based on the heteroscedastic noise model, which more accurately describes a natural raw image. This model is characterized by only two parameters, which are considered as unique fingerprint to identify camera models. The camera model identification problem is cast in the framework of hypothesis testing theory. In an ideal context where all model parameters are perfectly known, the likelihood ratio test (LRT) is presented and its performances are theoretically established. For a practical use, two generalized LRTs are designed to deal with unknown model parameters so that they can meet a prescribed false alarm probability while ensuring a high detection performance. Numerical results on simulated images and real natural raw images highlight the relevance of the proposed approach.

Conditional Density Models Integrating Fuzzy and Probabilistic Representations of Uncertainty

Abstract: markdown____ Conditional density estimation is an important problem in a variety of areas such as system identification, machine learning, artificial intelligence, empirical economics, macroeconomic analysis, quantitative finance and risk management. This work considers the general problem of conditional density estimation, i.e., estimating and predicting the density of a response variable as a function of covariates. The semi-parametric models proposed and developed in this work combine fuzzy and probabilistic representations of uncertainty, while making very few assumptions regarding the functional form of the response variable's density or changes of the functional form across the space of covariates. These models possess sufficient generalization power to approximate a non-standard density and the ability to describe the underlying process using simple linguistic descriptors despite the complexity and possible non-linearity of this process. These novel models are applied to real world quantitative finance and risk management problems by analyzing financial time-series data containing non-trivial statistical properties, such as fat tails, asymmetric distributions and changing variation over time.

Recursive Extended Least Squares Parameter Estimation for Wiener Nonlinear Systems with Moving Average Noises

Abstract: Many control algorithms are based on the mathematical models of dynamic systems. System identification is used to determine the structures and parameters of dynamic systems. Some identification algorithms (e.g., the least squares algorithm) can be applied to estimate the parameters of linear regressive systems or linear-parameter systems with white noise disturbances. This paper derives two recursive extended least squares parameter estimation algorithms for Wiener nonlinear systems with moving average noises based on over-parameterization models. The simulation results indicate that the proposed algorithms are effective.

Continuous Mixed $p$ -Norm Adaptive Algorithm for System Identification

Abstract: We propose a new adaptive filtering algorithm in system identification applications which is based on a continuous mixed $p$ -norm It enjoys the advantages of various error norms since it combines p-norms for $1 \leq p \leq 2$ The mixture is controlled by a continuous probability density-like function of $p$ which is assumed to be uniform in our derivations in this letter Two versions of the suggested algorithm are developed The robustness of the proposed algorithms against impulsive noise are demonstrated in a system identification simulation

Optimization of Perturbative PV MPPT Methods Through Online System Identification

Abstract: The maximum power point (MPP) tracking function allows one to extract the maximum power from the photovoltaic (PV) system in any operating condition. The perturb and observe technique is one of the mainly used algorithms aimed at this function. Its performances depend on the values of the two design parameters, which are the amplitude and the frequency of the perturbations that it applies in identifying the position of the MPP. In this paper, a method for the online optimization of the perturbation period is presented. It is based on the identification of the whole system, including the PV source and the dc/dc converter controlling it. The system impulse and frequency responses are evaluated by using the cross-correlation method. Such a technique, as well as the tracking algorithm, is effectively implemented by using a field-programmable gate array, and it is validated by means of simulation and experimental results.

Image-Based Visual Servoing of a 7-DOF Robot Manipulator Using an Adaptive Distributed Fuzzy PD Controller

Abstract: This paper is concerned with the design and implementation of a distributed proportional-derivative (PD) controller of a 7-degrees of freedom (DOF) robot manipulator using the Takagi-Sugeno (T-S) fuzzy framework. Existing machine learning approaches to visual servoing involve system identification of image and kinematic Jacobians. In contrast, the proposed approach actuates a control signal primarily as a function of the error and derivative of the error in the desired visual feature space. This approach leads to a significant reduction in the computational burden as compared to model-based approaches, as well as existing learning approaches to model inverse kinematics. The simplicity of the controller structure will make it attractive in industrial implementations where PD/PID type schemes are in common use. While the initial values of PD gain are learned with the help of model-based controller, an online adaptation scheme has been proposed that is capable of compensating for local uncertainties associated with the system and its environment. Rigorous experiments have been performed to show that visual servoing tasks such as reaching a static target and tracking of a moving target can be achieved using the proposed distributed PD controller. It is shown that the proposed adaptive scheme can dynamically tune the controller parameters during visual servoing, so as to improve its initial performance based on parameters obtained while mimicking the model-based controller. The proposed control scheme is applied and assessed in real-time experiments using an uncalibrated eye-in-hand robotic system with a 7-DOF PowerCube robot manipulator.

The advantages of directly identifying continuous-time transfer function models in practical applications

Abstract: The direct identification and estimation of continuous-time models from sampled data is now mature. This paper does not present any new methodology, nor does it compare the performance of existing methods. Its main aim is to discuss the advantages of direct, continuous-time model identification with the help of illustrative examples that are all based on real data from practical applications. Although the specific method of statistical parameter estimation is relatively unimportant in this regard, the latest and most sophisticated time domain identification algorithm that is freely available to the reader is used in these examples in order to ensure that the results reflect the best performance that can be achieved at this time by time-domain identification.

Identifying the dynamic model used by the KUKA LWR: A reverse engineering approach

Abstract: An approach is presented for the model identifi- cation of the so-called link dynamics used by the KUKA LWR- IV, a lightweight manipulator with elastic joints that is very popular in robotics research but for which a complete and reliable dynamic model is not yet publicly available. The control software interface of this robot provides numerical values of the link inertia matrix and the gravity vector at each configuration, together with link position and joint torque sensor data. Taking advantage of this information, a general procedure is set up for determining the structure and identifying the value of the relevant dynamic coefficients used by the manufacturer in the evaluation of these robot model terms. We call this a reverse engineering approach, because our main goal is to match the numerical data provided by the software interface, using a suitable symbolic model of the robot dynamics and the inertial and gravity coefficients that are being estimated. Only configuration-dependent terms are involved in this process, and thus static experiments are sufficient for this task. The main issues of dynamic model identification for robots with elastic joints are discussed in general, highlighting the pros and cons of the approach taken for this class of KUKA lightweight manip- ulators. The main identification results, including training and validation tests, are reported together with additional dynamic validation experiments that use the complete identified model and joint torque sensor data.