Interconnection of Uncertain Behavioral Systems for Robust Control
Abstract
This paper attempts to relate robust control and behavioral frameworks by incorporating structured uncertainty into the description of behavioral systems. Behavioral equations are expressed as linear fractional transformations (LFTs) on an uncertainty structure, and a method of interconnection is outlined. A method for obtaining input- output maps from LFT representations of behavioral systems is also presented. This extension of the behavioral framework is compatible with existing robust control methods, such as p analysis, which can be used to provide robustness tests in behaviors. A simple example is presented that illustrates some of the issues which arise in this extension. A major theme in robust control has been to supply the engineer with a theoretical and computational framework that handles a rich variety of modeling uncertainty so that physically motivated uncertainty descriptions can be treated in a natural manner. In particular, it has been important to provide computational tools that analyze systems with mixtures of unstructured uncertainties and possibly large numbers of uncertain real parameters. Behavioral models are in turn very natural when modeling physical systems from first principles, or when a large system is built up from subsystem models. While the final interconnected model used in a robust control design may have well-defined inputs and outputs, it is almost always the case that components are modeled in terms of mass, momentum, or energy balances or physical laws such as Newton's second law, Ohm's law, and so on. These components do not
Document Details
- Document Type
- Technical Report
- Publication Date
- Oct 15, 1993
- Accession Number
- ADA464346
Entities
People
- Fernando Paganini
- Raffaello D'andrea
Organizations
- California Institute of Technology