Orthogonal Lattice Modeling of Nonlinear Systems.
Abstract
The application of analysis lattice filters to the problem of determining the input to a system from observations of the system's output (i.e., deconvolution is discussed. Both linear and nonlinear systems are considered. Lattice filter modeling algorithms (Levinson and Schur) are presented. The theory of least squares inverse filters is reviewed. This leads to a discussion of the lattice filter, which in turn leads to the Generalized Lattice Theory. The Generalized Lattice Theory is then used to develop a nonlinear lattice structure. Simulations show that the nonlinear lattice is an effective inverse filter for both linear and nonlinear systems. Keywords: Autoregressive; Deconvolution; Lattice Filter; Nonlinear Lattice Filter; Kalman Filter; Least-squares Inverse Filter; Nonlinear Deconvolution. (Author)
Document Details
- Document Type
- Technical Report
- Publication Date
- Sep 01, 1986
- Accession Number
- ADA175147
Entities
People
- Scot L. Johnson
Organizations
- Naval Postgraduate School