Awomatica,Vol13.pp.457-476. Pergamon Press, 1977. Printed in Great Britain
Theory and Applications of
Self-Tuning Regulators*
K. J. ASTROM.* U, BORISSON.t L. LIUNGS
and B. WITTENMARK+
The theory of self-tuning regulators is reviewed. The performance of the self-tuning
regulatorsisillustrated by simulated examples and by applications on industrial processes.
Key, Word Index—Aduptive control; parameter estimation; process control: slFadjusting systems;
stochastic contol,
‘Summary —This paper reviews work on self-tuning regulators
The regulator algorhns, ther theory and industrial applic
cations are reviewed. The paper is expesitory—the major Hews
are covered but detailed analysis given elsewhere
1. INTRODUCTION
STOCHASTIC control theory has proved to be a
useful tool for the design of controllers for industrial
processes[1]J. In many practical applications itis,
however, dificult to determine the parameters of the
controller, since the dynamics of the process and its
disturbances are unknown. The parameters of the
process thus have to be estimated. For stationary
processes itis possible to determine the unknown
parameters through identification, The experiments
and their evaluations can, however, be rather time
consuming, It is thus desirable to have a regulator
which tunes its parameters on-line. The motivation
{or the work reviewed here has been to design such
regulators for control of industrial processes,
The class of regulators can be thought of as
composed of three parts: a parameter estimator, a
linear controller and a block which determines the
controller parameters from the estimated para-
‘meters, Within this structure there ate, of course,
many different possibilities, depending on the
control and estimation schemes used. The approach
isinno doubta very straightforward one. [tincludes,
‘Received 30 September 1976; revised 7 March 1977, An
cviginal version ofthis paper was presented at the 6h TPAC
Congress on onto Technolog the Sersceof Mean which was
held in Boston/Cambridge, MA. doring August 1975. The
published Proceedings of this IFAC Meeting may be ordered
fom: Instrument Society of America (ISA), 400 Stanek Steet,
Pittsbatgh, PA 15202. USA. or John Wiley. Balins Lane.
Chichester, Sussex, PO 19 IUD, U.K. This paper ws recon
‘mended for publication in revised form by assaciate editor H,
Kwakernaale
“Department of Automatic Control
Technology. Lund, Sweden.
{Granges Data, Nyby Brak, Sweden.
§Department of Electtical Enginecting, Linkoping Univesity,
Linkoping Sweden,
Lund Institute of
€8, application of the ‘certainty equivalence prin-
0. This value a* gives the minimum variance
controller when inserted in (3.13). Further analysis,
shows that the stability condition (3.7) is satisfied if
O