Adaptive nonlinear system identification. The Volterra and Wiener model approaches. (Q2458636): Difference between revisions

In this book, the author presents simple, concise, easy-to-understand methods for identifying nonlinear systems using adaptive filter algorithms well known for linear systems identification. The book focuses on the Volterra and Wiener models for nonlinear systems, but there are other nonlinear models as well. The focus here is on one-dimensional signal processing. However, much of the material presented here can be extended to two- or multi-dimensional signal processing as well. This book is not exhaustive of all the methods of nonlinear adaptive system identification. It is another contribution to the current literature on the subject. The book will be useful for graduate students, engineers and researchers in the area of the nonlinear systems identification and adaptive signal processing. The book is organized as follows. There are three parts. Part 1 consists of chapters 1 through 5. These contain some useful background material. Part 2 describes the different gradient-type algorithms and consists of chapters 6 through 9. Part 3, which consists only of chapter 10, describes the recursive least-squares-type algorithms. Chapter 11 has the conclusions. Chapter 1 introduces the definition of nonlinear systems. Chapter 2 introduces polynomial modeling for nonlinear systems. In chapter 3, both Volterra and Wiener models for nonlinear systems are introduced. Chapter 4 reviews the methods used for system identification of nonlinear systems. Chapter 5 reviews the basic concepts of adaptive filter algorithms. Stochastic gradient-type adaptive system identification methods based on Volterra model are presented in chapter 6. In chapters 7 and 8, the algorithms for nonlinear adaptive system identification of second- and third-order Wiener models respectively are presented. Chapter 9 extends this to other related stochastic-gradient-type adaptive algorithms. In chapter 10, recursive-least-squares-type algorithms for Wiener model of nonlinear system identification are described. Chapter 11 contains the summary and conclusions.