Preface
these methods
to PID
control.
At
the same time,
it seemed
natural to
bring
PID
control closer
to the mainstream
ideas in control. When working with
in-
dustrial auto-tuners,
we also saw
a
great
need to include diagnostics in the
controller, because
it is no use
to
tune a controller
if
the
process
has
severe
malfunctions.
The
present
book,
Aduanced
PID
Control, is the result
of this
effort.
With
a
total
of
13
chapters,
this
new
book substantially expands on some
of
the topics covered
in the
previous
versions
and
provides
several new
chapters
that
deal
with controller design,
feedforward design,
replacement
of the Ziegler-
Nichols tuning
rules,
predictive
control,
loop
and
performanee
asqeqqrnnnr
?-i
interaction. At
this
point
in
our book
trilogy, we assume that the
reader rb
highly
familiar
with
control theory.
Our
research has
given
a deeper understanding
of the trade-offs between
load disturbance attenuation,
injection of measurement noise, and
set-point
response. We
have
also been
able
to
answer
questions
like:
Should a controller
be tuned
for response to
load
disturbances
or set
points?
What
information is
required to design
a
PID
controller?
When can
derivative action
give
signif-
icant improvements? When
are
more
complicated
controllers
justified?
When
is
it
justified
to develop
more accurate
process
models? With the
knowledge
developed,
it
is
now
possible
to design auto-tuners
that can
make
these
assess-
ments autonomously.
In
addition,
we have developed
new
simple
methods
for
designing
PID controllers.
As
an example
of the
insight
gained
we can
mention that
control
theory
tells that
it is not
necessary
to
make a compromise
between tuning
for load
disturbance
response
and set-point
response. Both requirements can
be
satis-
fied
by
using a controller
with two
degrees of freedom, which
combines
feedback
and
feedforward. The
feedback
gains
should be chosen
to
satisfy
requirements
on
disturbance
attenuation
and
robustness. The
desired
response to
set-point
changes can then be
obtained by
proper
use of
feedforward. Set-point
weighting
is
a
simple
form of
feedforward
for PID control.
In
some cases,
it is
justified
to
use
more
elaborate
feedforward.
For
this
reason,
we
have included a
chapter
on controller
design
and another
chapter on feedforward in the
new book.
The
robustness
analysis
also
shows
the
advantage
of having low controller
gain
at
high frequency, high
frequency
roll-off. This
can be accomplished
by
flltering
the
process
output
by a second order
filter. Based on
the
insight
ob-
tained,
we recommend extended
use
of
set-point weighting or
more advanced
feedforward. We also
recommend that the
process
output
is
fiItered
using
a
second
order
filter.
We
would
like
to
thank many
people
who have
given
knowledge,
insight,
and
inspiration.
Our
interest in
PID
control was
inspired by
Axel
Westrenius
and
Mike
Somerville
of Eurotherm
in
the early
1980s.
We have learned
much
from working with students;
particular
thanks are due
to Lars
Gdran
Elfgren
(Eurotherm), Giiran Grtinhammar
(LTH),
Ari
Ingimundarson (UPC), Oskar
Nordin
(Volvo),
Helene Panagopoulos
(Volvo),
Per Persson
(Volvo),
Mikael Pe-
tersson
(ABB),
Ola
Slattke
(ABB),
and
Anders Wall6n
(Ericsson Mobile Plat-
forms), who continue
to
give
us
valuable
insight even if they are now
pursuing
careers
in
industry.
We
are very
grateful
to Sune
Larsson
and
Lars Beath,
formerly
of
NAF
vi
评论0
最新资源