Note: This content is accessible to all versions of every browser. However, this browser does not seem to support current Web standards, preventing the display of our site's design details.


Nonlinear Structures in Instrumentation, Modeling, and Data Analysis


R. Pearson

Daettwil, Switzerland, ABB Corporate Research Center

Nonlinear, discrete-time dynamic model structures arise naturally in many industrially-important process monitoring and control problems. To illustrate this point, this talk begins with brief discussions of three examples: modeling the changes in sensor dynamics caused by fouling, gray-box modeling for nonlinear process control (e.g., nonlinear model predictive control), and the design of data cleaning filters for removing outliers from measurement data. In all of these problems, we are immediately faced with the question of how to choose a nonlinear dynamic model structure, either to describe observed dynamic behavior or to accomplish a nonlinear filtering task. This talk gives a brief overview of some of the major structural options we have to choose from (e.g., nonlinear FIR or moving average models vs. output-affine models vs. nonlinear autoregressive models) and some simple qualitative criteria that can be helpful in making this choice (e.g., in model-building, how strongly does the dynamic character change with different inputs? in digital filtering, do we want to impose scale-invariance or other behavioral constraints?). Taken together, the examples and results presented illustrate some of the important and often surprising ways nonlinear dynamic models do and do not differ from "similar" linear structures from which they are derived.


Type of Publication:

(05)Plenary/Invited/Honorary Lecture

File Download:

Request a copy of this publication.
(Uses JavaScript)
% No recipe for automatically generating a BibTex entry for (05)Plenary/Invited/Honorary Lecture
Permanent link