Identification results are presented from both deterministic (least squares) and stochastic perspectives
The identification of Volterra series is addressed from a new perspective, bringing to light many new and significant results
Both direct synthesis techniques as well as more general, implicit formulations are discussed
This book covers recent results in the analysis, identification and control of systems described by Volterra models. Topics covered include: qualitative behavior of finite Volterra models compared and contrasted with other nonlinear model classes, structural restrictions and extensions to Volterra model class, least squares and stochastic identification approaches, model inversion issues, and direct synthesis and model predictive control design, guidelines for practical applications. Examples are drawn from Chemical, Biological and Electrical Engineering. The book is suitable as a text for a graduate control course, or as a reference for both research and practice.|Much has been written about the general difficulty of developing the models required for model-based control of processes whose dynamics exhibit signif icant nonlinearity (for further discussion and references, see Chapter 1). In fact, the development ofthese models stands as a significant practical imped iment to widespread industrial application oftechniques like nonlinear model predictive control (NMPC), whoselinear counterpart has profoundly changed industrial practice. One ofthe reasons for this difficulty lies in the enormous variety of "nonlinear models," different classes of which can be less similar to each other than they are to the class of linear models. Consequently, it is a practical necessity to restrict consideration to one or a few specific nonlinear model classes if we are to succeed in developing, understanding, and using nonlinear models as a basis for practical control schemes. Because they repre sent a highly structured extension ofthe class oflinear finite impulse response (FIR) models on which industrially popular linear MPC implementations are based, this book is devoted to the class of discrete-time Volterra models and a fewother, closelyrelated, nonlinear model classes. The objective ofthis book is to provide a useful reference for researchers in the field of process control and closely related areas, collecting a reasonably wide variety of results that may be found in different parts of the large literature that exists on the gen eral topics of process control, nonlinear systems theory, statistical time-series models, biomedical engineering, and digital signal processing, among others.
1. Introduction.- 2. Qualitative Behavior.- 3. Restrietions & Extensions.- 4. Determination of Volterra Model Parameters.- 5. Practical Considerations in Volterra Model Identification.- 6. Model-Based Controller Synthesis.- 7. Advanced Direct Synthesis Controller Design.- 8. Model Predictive Control Using Volterra Series.- 9. Application Case Studies.- 10. Summary.
Much has been written about the general difficulty of developing the models required for model-based control of processes whose dynamics exhibit signif icant nonlinearity (for further discussion and references, see Chapter 1). In fact, the development ofthese models stands as a significant practical imped iment to widespread industrial application oftechniques like nonlinear model predictive control (NMPC), whoselinear counterpart has profoundly changed industrial practice. One ofthe reasons for this difficulty lies in the enormous variety of "nonlinear models," different classes of which can be less similar to each other than they are to the class of linear models. Consequently, it is a practical necessity to restrict consideration to one or a few specific nonlinear model classes if we are to succeed in developing, understanding, and using nonlinear models as a basis for practical control schemes. Because they repre sent a highly structured extension ofthe class oflinear finite impulse response (FIR) models on which industrially popular linear MPC implementations are based, this book is devoted to the class of discrete-time Volterra models and a fewother, closelyrelated, nonlinear model classes. The objective ofthis book is to provide a useful reference for researchers in the field of process control and closely related areas, collecting a reasonably wide variety of results that may be found in different parts of the large literature that exists on the gen eral topics of process control, nonlinear systems theory, statistical time-series models, biomedical engineering, and digital signal processing, among others.
1. Introduction.- 2. Qualitative Behavior.- 3. Restrietions& Extensions.- 4. Determination of Volterra Model Parameters.- 5. Practical Considerations in Volterra Model Identification.- 6. Model-Based Controller Synthesis.- 7. Advanced Direct Synthesis Controller Design.- 8. Model Predictive Control Using Volterra Series.- 9. Application Case Studies.- 10. Summary.
Inhaltsverzeichnis
1. Introduction.- 2. Qualitative Behavior.- 3. Restrietions & Extensions.- 4. Determination of Volterra Model Parameters.- 5. Practical Considerations in Volterra Model Identification.- 6. Model-Based Controller Synthesis.- 7. Advanced Direct Synthesis Controller Design.- 8. Model Predictive Control Using Volterra Series.- 9. Application Case Studies.- 10. Summary.
Klappentext
Much has been written about the general difficulty of developing the models required for model-based control of processes whose dynamics exhibit signif icant nonlinearity (for further discussion and references, see Chapter 1). In fact, the development ofthese models stands as a significant practical imped iment to widespread industrial application oftechniques like nonlinear model predictive control (NMPC), whoselinear counterpart has profoundly changed industrial practice. One ofthe reasons for this difficulty lies in the enormous variety of "nonlinear models," different classes of which can be less similar to each other than they are to the class of linear models. Consequently, it is a practical necessity to restrict consideration to one or a few specific nonlinear model classes if we are to succeed in developing, understanding, and using nonlinear models as a basis for practical control schemes. Because they repre sent a highly structured extension ofthe class oflinear finite impulse response (FIR) models on which industrially popular linear MPC implementations are based, this book is devoted to the class of discrete-time Volterra models and a fewother, closelyrelated, nonlinear model classes. The objective ofthis book is to provide a useful reference for researchers in the field of process control and closely related areas, collecting a reasonably wide variety of results that may be found in different parts of the large literature that exists on the gen eral topics of process control, nonlinear systems theory, statistical time-series models, biomedical engineering, and digital signal processing, among others.
Identification results are presented from both deterministic (least squares) and stochastic perspectives
The identification of Volterra series is addressed from a new perspective, bringing to light many new and significant results
Both direct synthesis techniques as well as more general, implicit formulations are discussed
