/o-1245369/n/identification-of-dynamical-systems)
Precise dynamic models of processes are required for many applications, ranging from control engineering to natural sciences and economics. Frequently, such precise models cannot be derived by theoretical considerations only. Therefore, they must be determined experimentally. The book gives an introduction to the determination of dynamic models based on measurements taken at the process, which is known as system identification or process identification respectively. Both offline and online methods are treated, i.e. methods that post-process the measured data as well as methods that provide models during the measurement. The book is application oriented and most methods covered have been used successfully in practical applications for many different processes.
After a short introduction into the required basic relations for continuous time and discrete time linear systems, the focus is first on the identification of non-parametric models with continuous time signals. Here, methods are presented that allow a simple and quick inspection based on step and impulse responses. This is followed by methods for the determination of the non-parametric frequency response of systems, based on the Fourier transform and correlation analysis. Then, the parameter estimation for parametric models is presented with a focus on the method of Least Squares and some of its most prominent modifications. Issues such as parameter estimation for time-variant processes, parameter estimation in closed-loop, parameter estimation for differential equation models of continuous time processes and efficient implementations of the algorithms are discussed. The different methods are compared by using a real three mass oscillator process.
Some methods for nonlinear system identification are considered such as parameter estimation for Volterra, Hammerstein and Wiener models, the Extended Kalman filter, neural networks and look-up tables. The application of the methods is illustrated on examples with real measurements ranging from hydraulic systems to combustion engines. Real experimental data are also given on a CD-ROM, allowing to test many of the methods presented in this book with real measurements. The book is dedicated to students and practising engineers in research and development, design and manufacturing. TOC:Introduction .- Mathematical Models of Linear Dynamic Systems and Stochastic Signals.- Identification with Non-Parametric Models - Continuous Time Signals.- Identification with Non-Parametric Models - Discrete Time Signals.- Identification with Parametric Models - Discrete Time Signals.- Identification with Parametric Models - Continuous Time Signals.- Identification of Multi-Variable Systems.- Identification of Non-Linear Systems.- Miscellaneous Issues.- Appendix
After a short introduction into the required basic relations for continuous time and discrete time linear systems, the focus is first on the identification of non-parametric models with continuous time signals. Here, methods are presented that allow a simple and quick inspection based on step and impulse responses. This is followed by methods for the determination of the non-parametric frequency response of systems, based on the Fourier transform and correlation analysis. Then, the parameter estimation for parametric models is presented with a focus on the method of Least Squares and some of its most prominent modifications. Issues such as parameter estimation for time-variant processes, parameter estimation in closed-loop, parameter estimation for differential equation models of continuous time processes and efficient implementations of the algorithms are discussed. The different methods are compared by using a real three mass oscillator process.
Some methods for nonlinear system identification are considered such as parameter estimation for Volterra, Hammerstein and Wiener models, the Extended Kalman filter, neural networks and look-up tables. The application of the methods is illustrated on examples with real measurements ranging from hydraulic systems to combustion engines. Real experimental data are also given on a CD-ROM, allowing to test many of the methods presented in this book with real measurements. The book is dedicated to students and practising engineers in research and development, design and manufacturing. TOC:Introduction .- Mathematical Models of Linear Dynamic Systems and Stochastic Signals.- Identification with Non-Parametric Models - Continuous Time Signals.- Identification with Non-Parametric Models - Discrete Time Signals.- Identification with Parametric Models - Discrete Time Signals.- Identification with Parametric Models - Continuous Time Signals.- Identification of Multi-Variable Systems.- Identification of Non-Linear Systems.- Miscellaneous Issues.- Appendix
