Synthesis of
multivariate control systems by objects with inexact data
Yunicheva
N.R.
e-mail: naduni@ mail.ru
Institute of Informatics and Control Problems,
Almaty, Kazakhstan
For the decision of tasks of the analysis and synthesis, research of
quality indicators of intellectual control systems interval and indistinct
methods [1] are used.
Application of the given methods is caused by
that discrepancies of the data in parameters of object of management or,
otherwise, uncertainty have a statistical property.
In given article the decision of task of
parametrical synthesis for multivariate intellectual
control systems of objects with the inexact data is carried out with the
help of a method [2] and methods of the interval analysis. The computing
algorithm in C ++ is developed.
1
Statement of task
Let's assume,
that the mathematical model of multivariate uncertain object of control looks
as follows:
, (1)
where - a n-dimensional vector
of conditions, - a m-dimensional
vector of control actions, – interval matrixes of control object.
The
purpose of synthesis task we shall count definition of an interval matrix in the equation
of feedback
, (2)
providing reception of desirable
dynamic properties in the closed control system it will be presented in the
following kind:
, (3)
.where – an interval matrix of the closed control system.
The interval characteristic polynomial
of the closed system will look like:
, (4)
where - interval numbers of
a characteristic polynomial.
The algorithm of
the task decision in view of synthesis will consist of the following steps:
Step 1:
Decision of
parametrical synthesis task for multivariate stationary object of the
management belonging to family (1).
Before directly
to proceed to the decision of a task in view of synthesis, from family of
mathematical models (1) we shall allocate mathematical model with the fixed
values of elements of matrixes , i.e. mathematical
model of the following kind:
(5)
where and – constant matrixes. For model of a kind (5) equation of a
feedback will be presented as follows:
(6)
where K – the matrix of adjusted parameters, which
definition will answer the decision of a problem(task) of synthesis for
stationary control object, carried out
on a technique suggested in [3].
Step 2:
Decision of
parametrical synthesis task for multivariate object of management on the basis
of a method of common parameter and the matrix found above K as initial approximation the decision of
synthesis task of control we shall increase the first column of a matrix K
on the common parameter β:
or
Then instead of
the equation (2) we shall write down:
(7)
The
characteristic polynomial of the closed control system with the account (7)
will be written down in the following kind:
(8)
β. also it should be equal to a desirable characteristic polynomial (4).
Equating factors at identical degrees λ expressions (9) and (4), we shall
receive system of the linear algebraic interval equations for definition of
parameter β.
Let's consider
the equation of the received system which looks like:
, (9)
where - the interval sizes dependent on
elements of matrixes and K .
By formal
disclosing brackets and reduction similar composed we shall receive expression:
(10)
According to the
subdistributive law of interval mathematics for set β we have:
,
(11)
The parameter will be found from
expression:
, (12)
The found value is substituted in:
,
(13)
Thus, received in
view of found factors of the
characteristic equation of the closed system will be included in factors of the
desirable characteristic equation. The system of the equations (13) according
to a synthesis algorithm will be transformed to a kind:
, (14)
The opportunity of such
transformation is determined with the help of operation of accommodation by
summation [3].
The decision of
the received system (15) is given with the following theorem: the valid
parameter is the decision íåâûðîæäåííîé systems of the linear algebraic equations in only case when it represents the decision of system of the
equations:
, (15)
also satisfies to restrictions:
,
(16)
where the centers and lengths of interval
numbers .
From (16) sequences of
parameters found on a condition it is chosen the minimal
value of the adjusted parameter.
1.
Jolen L., Kiefer Ì., Didri Î. Walter E.. The applied interval analysis. M.: Institute of computer
researches. 2007. – 467p.
2.
Ashimov
A.A., Syzdykov D.Zh. Identification by the common parameter method: Reference
book on the theory of automatic control / Edited by Krasovsky A.A. –
M.:Science, 1987. –P.263-271.
3.
Khlebalin
N.A. Modal Control of Plants with Uncertain Interval Parameters, in: Proc.
Intern. Workshop «Control System Syntesis: Theory and Application»,
Novosibirsk, 1991. -P. 168-173.