Adaptive Nonlinear System Identification: The Volterra and by Tokunbo Ogunfunmi

By Tokunbo Ogunfunmi

Makes a speciality of approach id functions of the adaptive equipment awarded. yet which may even be utilized to different purposes of adaptive nonlinear procedures. Covers contemporary study leads to the world of adaptive nonlinear process identity from the authors and different researchers within the box.

Show description

Read or Download Adaptive Nonlinear System Identification: The Volterra and Wiener Model Approaches (Signals and Communication Technology) PDF

Similar linear programming books

Techniques in Variational Analysis

Variational arguments are classical suggestions whose use could be traced again to the early improvement of the calculus of diversifications and extra. Rooted within the actual precept of least motion, they've got vast functions in assorted fields. This ebook offers a concise account of the basic instruments of infinite-dimensional first-order variational research.

Introduction to Nonlinear Physics

This textbook offers an advent to the hot technological know-how of nonlinear physics for complicated undergraduates, starting graduate scholars, and researchers coming into the sector. The chapters, by means of pioneers and specialists within the box, percentage a unified viewpoint. Nonlinear technological know-how built out of the expanding skill to enquire and study platforms for which results usually are not easily linear features in their motives; it truly is linked to such recognized code phrases as chaos, fractals, trend formation, solitons, mobile automata, and intricate structures.

Iterative methods for optimization

This ebook provides a gently chosen workforce of tools for unconstrained and certain restricted optimization difficulties and analyzes them intensive either theoretically and algorithmically. It makes a speciality of readability in algorithmic description and research instead of generality, and whereas it offers tips to the literature for the main common theoretical effects and strong software program, the writer thinks it truly is extra very important that readers have an entire knowing of targeted instances that express crucial principles.

Variational Methods for Structural Optimization

In fresh many years, it has turn into attainable to show the layout method into computing device algorithms. by means of making use of diversified machine orientated equipment the topology and form of constructions will be optimized and hence designs systematically stronger. those percentages have prompted an curiosity within the mathematical foundations of structural optimization.

Extra info for Adaptive Nonlinear System Identification: The Volterra and Wiener Model Approaches (Signals and Communication Technology)

Sample text

Orthogonal Tchebyshev Polynomials Tchebyshev’s differential equation arises as a special case in the SturmLiouville boundary value problem which is defined as (1 − x 2 ) y ''' − xy ' + n 2 y = 0 where n is a real number. The solutions of this differential equation are referred to as Tchebyshev’s functions of degree n . For non-negative n, n = 0,1, 2, 3,........ , Tchebyshev’s functions are referred to as Tchebyshev’s polynomials, T n ( x) . Tchebyshev’s polynomials T n ( x) can be expressed as (Efunda 2006): T n ( x) = 1 − x2 dn n− 1 (1 − x 2 ) 2 where n = 0,1, 2,3......

It is directly based on the Volterra series called the Volterra model. The advantage of the Volterra model is that there is little or no preprocessing needed before the adaptation. But because of the statistically nonorthogonal nature of the Volterra space spanned by the Volterra series components, it is necessary to perform the Gram-Schmidt/modified Gram-Schmidt procedure 20 Chapter 2 or QR decomposition method to orthogonalize the inputs. This orthogonalization procedure is crucial especially for the nonlinear LMS-type algorithms and also for the nonlinear RLS-type recursive Volterra adaptive algorithms.

In general, the positive powers of x(n) are x ( n), x 2 ( n), x 3 (n), x 4 (n), x 5 (n),...... Let x(n) and y(n) represent the input and output signals, respectively. For a linear causal system, the output signal y(n) can be expanded as the linear combination of M-memory input signal x(n) as y(n) = c0x(n) + c1x(n-1) + c2x(n-2) + c3x(n-3) + ... 1) where ck are the filter coefficients representing the linear causal system. , x(n-M+1)}. Taking advantage of this input orthogonal property, a lot of linear adaptive algorithms have been developed (Widrow 1985, Haykin 1996, Diniz 2002, Sayed 2003).

Download PDF sample

Rated 4.38 of 5 – based on 41 votes