Uppsala universitet
Optimal and Adaptive Feedforward Regulators

Mikael Sternad

Doctoral dissertation at Uppsala University
ISBN 91-7900-219-6
May 1987, 215 pp.

Thesis in PDF

Paper copies of the thesis can be obtained from Mikael Sternad, Signals and Systems Group, Uppsala University, Box 534, SE-75121 Uppsala, Sweden.

When disturbances can be measured, a large improvement of disturbance rejection becomes possible, compared with the use of output feedback only.

Polynomial LQG design methods for feedforward and combined feedforward-feedback control of discrete time single input systems are presented. These regulators attain optimal disturbance rejection in cases where complete cancellation of disturbances is impossible, for example when systems are non-minimum phase. The optimal regulator structure is discussed. With the help of this structure, the feedforward control quality is independent of the choice of feedback. The methods handle deterministic as well as stochastic disturbances. Disturbance measurement signals affected by the input can be utilized.

An application to load feedforward power frequency control of hydro power stations is studied.

A close correspondence is shown to exist between feedforward control and input estimation problems.

Two adaptive feedback control algorithms are extended with adaptive feedforward: An LQG self-tuner and explicit criterion minimization, which both converge to the optimal controller. Their performance is tested and compared to minimum variance and extended minimum variance self-tuners with feedforward terms. Both the LQG self-tuner and explicit criterion minimization attain very good control behaviour. With the help of some simple safeguards, their performance is robust under a wide range of conditions.

Feedforward control, Linear quadratic control, Disturbance rejection, Disturbance decoupling, Power systems control, Deconvolution, Adaptive control Optimal control.

Table of Contents:
  1. Introduction
  2. Five approaches to feedforward and disturbance decoupling control
  3. Feedforward LQG regulator design
  4. Examples and a case study
  5. Alternative adaptive control strategies
  6. Performance of the adaptive regulators
  7. Conclusions

| Audio research | Polynomial methods | Main entry in list of publications |
This material is presented to ensure timely dissemination of scholarly and technical work. Copyright and all rights therein are retained by authors. All persons copying this information are expected to adhere to the terms and constraints invoked by each authors copyright. This work may not be reposted without the explicit permission of the copyright holders.