Professional Documents
Culture Documents
Pal 1977
Pal 1977
Pal 1977
Energy loss per cycle Tempering temperature X590 480 370 260 150 25
kJ/m 3 Magnetising force kA/m 6-7 9-2 10-8 130 15-4 19-6
12 Flux density T 1-24 0-93 0-8 0-7 0-64 0-57
Magnetising force kA/m 8-6 10-6 120 140 16-5 20-6
18 Flux density T 1-4 112 0-97 0-87 0-78 0-68
Magnetising force kA/m 12-6 130 13-8 150 16 7 22-0
24 Flux density T 1-7 1-26 11 0-99 0-89 0-76
OPTIMAL AND SUBOPTIMAL CONTROL OF index / of eqn. 3 of the paper; the general practice is to determine
DUAL-EXCITED SYNCHRONOUS GENERATORS suitable values of <2, and Q2 by trial and error, where intuition again
plays a very important role. Because of the uncertainty in choosing
In the introduction to Paper 7764 P [Proc IEE, 1976,123, (10), these matrices, the optimal controllers designed are not necessarily the
pp. 989-992], Daniels and Lee claim that most previous attempts to best. Also, in certain situations the linear optimal controller results in
improve the transient performance of the dual-excited synchronous decreased stability limits.0
machines have been based on an intuitive approach. They have then
used linear optimal-excitation control to supposedly overcome this 18th April 1977 O.P. MALIK
drawback, as optimal control theory provides a more systematic Department of Electrical Engineering
approach. They have, however, chosen to ignore the previous University of Calgary
workA>B relating to the application of optimal control to dual-excited Calgary, Canada T2N IN4
synchronous generators. In fact, Reference A and B employ nonlinear
optimal-control techniques for the improvement of the transient
performance of the dual-excited machines, and have thus gone further The list of References given in the paper was never intended to be a
than the work reported in this paper. complete bibliography on the subject of either the dual-excited
In their paper, the authors further claim that the system performance machine or control methods for such a machine. The main object of
with linear control is inferior to that obtained from the suboptimal the paper was to compare the performance of such a machine under
control described by eqns. 10 and 11 of the paper. The authors do not the action of two closed-loop excitation control systems: the optimal
give any systematic basis on which the suboptimal feedback controls controller based on a linearised model and the so-called suboptimal
and the performance index of eqns. 10 to 12 are derived. It would controller based on a nonlinear model and obtained by function
seem to me that the structure of their suboptimal control is based on minimisation using dynamic-sensitivity analysis. In this context,
an intuitive approach that is no different from that employed in References A and B of Prof. Malik's correspondence item do not
determining the additional signals for the 'conventional control'. appear to be relevant, since both papers referred to seem to derive
Maybe the authors would like to either comment on this or give a open-loop controllers. Analytical studies of such open-loop controllers
more systematic basis for the suboptimal control structure employed may well go further academically than the work reported in our
by them. paper, but are, we suggest, very difficult to implement on real physical
It should be pointed out further that as there are no guidelines systems, a point confirmed by comments in Reference C of the corres-
available in choosing the matrices Qy and Q2 in the performance pondence. Incidentally, none of the three References quoted appears
792 PROC. IEE, Vol. 124, No. 9, SEPTEMBER 1977