Undesirable Performance Characteristics of Existing Model-Reference Adaptive Control Algorithms.

Abstract

The full paper will describe recent theoretical results, backed up by digital computer simulation studies, that relate to the performance of several adaptive control algorithms that are based on model reference techniques. Special emphasis was placed upon the transient performance of these algorithms and the implications upon the bandwidth of the closed-loop adaptive system. The conclusions are as follows: (1) During the transient phase of the adaptation procedure, the control signal is characterized by excessive high frequency content. Novel analytical based upon stability theory can be used to predict this high frequency oscillatory behavior, which depends both upon the amplitude and frequency of the reference input. (2) During the transient adaptation phase the excessive control loop bandwidth is detrimental to system performance, because it can excite unmodeled high-frequency dynamics and lead to instability. (3) Similar effects can occur in the presence of stochastic inputs (plant noise and measurement noise). The analytical techniques that have been developed are constructive, so that modifications to the existing algorithms are suggested to overcome the practical shortcomings of the existing algorithms. Full details will be given in the conference paper. (Author)

Open PDF

Document Details

Document Type
Technical Report
Publication Date
Mar 15, 1981
Accession Number
ADA098099

Entities

People

  • C. Rohrs
  • G. Stein
  • L. Valavani
  • M. Athans

Organizations

  • Massachusetts Institute of Technology

Tags

Communities of Interest

  • Materials and Manufacturing Processes
  • Sensors

DTIC Thesaurus Topics

  • Adaptive Systems
  • Algorithms
  • Amplitude
  • Bandwidth
  • Closed Loop Systems
  • Computer Simulations
  • Control Systems
  • Differential Equations
  • Digital Computers
  • Dynamics
  • Equations
  • Frequency
  • Instability
  • Measurement
  • Nonlinear Differential Equations
  • Regulators
  • Simulations

Readers

  • Adaptive Control and Estimation with Uncertainty in Dynamic Systems.
  • Computational Modeling and Simulation
  • Control Systems Engineering.