Controller Design for Linear Stochastic Systems with Uncertain Parameters.

Abstract

The design of optimal controllers for linear stochastic systems requires an accurate description of the system. However, the construction of an accurate model of real systems is often not possible. These inaccuracies can stem from the fact that the adopted linear model may be only a first-order approximation of a nonlinear system. Also, there may be actual uncertainty in the parameters of the real system. This type of uncertainty can arise, for instance, if one wishes to design a single type of controller for a large number of similar systems, when, for example, the system is mass produced and it is impractical to tune each controller to each system. Another situation where this type of uncertainty can arise is when the parameters of a single system vary slowly over long periods of time, perhaps due to wear or changes in the environment. These uncertainties, in this thesis, are grouped into a vector of parameter. Several ways of handling these uncertainties in the design process are considered. Keywords: Minimax technique; Two dimensional.

Open PDF

Document Details

Document Type
Technical Report
Publication Date
Mar 01, 1985
Accession Number
ADA161428

Entities

People

  • Paul Howard Mcdowell

Organizations

  • University of Illinois Urbana–Champaign

Tags

Communities of Interest

  • Air Platforms
  • Materials and Manufacturing Processes

DTIC Thesaurus Topics

  • Algorithms
  • Closed Loop Systems
  • Control Systems
  • Convex Sets
  • Covariance
  • Discrete Distribution
  • Equations
  • Feedback
  • Illinois
  • Order Statistics
  • Quadratic Equations
  • Regulators
  • Statistics
  • Stochastic Control
  • Stochastic Processes
  • Transfer Functions
  • Two Dimensional

Fields of Study

  • Mathematics

Readers

  • Computational Modeling and Simulation
  • Operations Research
  • Robotics and Automation.