Development of a System-Grounded Criterion of Optimal Control
I. Lutsenko
Enterprises as structures combining great number of
controlled systems are created to meet their organizers
requirements in the form of a profit at the expense of the
possibility to use a generated cash flow.
It is quite natural that enterprise organizers seek to maximize their potentialities
opportunities resulted from the earned profit. Accordingly, they seek to maximize
receipt sizes within the certain time period by optimizing an enterprises controlled
systems.
But at present the operation mode of most of controlled systems is rather far from
being optimal. On the one hand it is caused by the nature of design that does not
allow of optimal control; on the other hand it is due to absence of an efficiency
index that can be used as a criterion of optimal (the best for production structure
organizers) control.
Here is the way the problem of optimal control criterion was treated last century:
Engineers, researchers, economists and designers are continuously suggesting
new universal, accurate and clear objective functions. One of the authors
succeeded in gathering over 100 criteria of optimizing division processes. in 1967.
After their classification it has become clear that there is no universal criterion and
choosing a process optimization or effectiveness criterion is quite a complicated
task [1].
For example, according to the process-oriented standard ISO series 9000:2000
efficiency is defined as gained results-consumed resources relationship [2].
As is seen, the situation has not changed. Even that fundamental document only
defines efficiency at the conceptual level.
RESOURCE USE EFFICIENCY FACTOR
(excerpts from the [Link])
Any process as well as a continuous process can be presented as consequence of
operations (fig.1).
Figure 1: Model of the process
The signal sensor - this is a motion detection signals of technological products (fig.
1-3).
1
( ) rq t
- the signal recording the motion of a special product
2
( ) rq t
- the signal recording movement of energy products
pq(t) - motion detection signal of the finished product
Figure 2: Registration model of operations
Input product cost for the studied technological process at a certain moment of
time is determined in the following way:
where i - input product identifier;
I - number of significant input products;
( )
i
rq t
- account parameter of an input product;
i
RS
- cost estimation of an account parameter of an input product;
re(t) - current value of a cumulative cost estimate of input products.
The above signal recording re(t)=re1(t)+re2(t).
By-turn cost estimation of output products of the studied process at certain moment
is determined as follows:
where j - output product identifier;
J - number of significant input products;
( )
j
pq t
- account parameter of an input product;
( )
j
PS t
- cost estimation of an account parameter of an input product;
pe(t) - current value of a cumulative cost estimate of input products.
For the constants
1
RS
,
2
RS
and PS were taken the following values:
1
RS
=1;
2
RS
=0.3/1000; PS=3.5
re(t) - the total number of registration for the input products at comparable cost values
pe(t) - registration number of finished products at comparable cost values
Figure 3: Registration model of operation research at comparable cost values
Process operation, which is reduced to a comparable value can also lead to its
borders.
Time moment corresponding to the complete transformation of input products of
an agreed operation is defined as moment of operations physical completion
(MOPC). MOPC is marked as f
t
on the diagrams. In turn, any process can be
presented as a reduced vector operation (fig.4).
re(t) - current value of a cumulative cost estimate of input products
pe(t) - current value of a cumulative cost estimate of input products
Figure 4: The model, which shows the signals to the boundaries of technological
operations
Consequently, it is important to determine the amplitudes of ES input and output
signals.
Input product cost for the studied technological process at a certain moment of
time is determined in the following way:
where re(t) - current value of a cumulative cost estimate of input products;
s
t
- the moment of process beginning;
pe(t) - current value of a cumulative cost estimate of input products;
f
t
- moment of operations physical completion.
RE and PE are the amplitudes of signals reflecting integral values of registered
characteristics and, therefore, of input and output products in comparable values.
An impulse signal RE is correlated with the moment of process beginning (
s
t
)
and provides re(t) function forming, and an impulse signal PE is correlated with
MOPC to form the pe(t) function.
Definition of absolute potential effect and resource capacity of technological
operation allows to obtain resource use efficiency factor as a ratio
A
E
R
.
Substituting
A
and
R
for corresponding integral expressions we get [3-6]
0 0 0
0 0 0 0 0
( ) ( )
( ) ( )
d
l
l
t
t
t
t
pe d re dv d d
E
re dv dv pe dv dv dv
_
1
1
1
]
,
1
_
1
1
,
]
,
[ ]
0 0 0 0 0
; ; ;
l d d
t t t t
;
0
[ , ]
d
v t t
;
[ , ]
l d
t t
;
0
[ , ]
l
t t
.
Resource use efficiency factor for the given operation we get in the form of
analytical expression
where op
T
- time of technological operations;
1
p
T
.
Conclusion. Resource use efficiency factor which can be used as a criterion of
optimization, showing control that guarantees maximization of financial potential
of controlling system has been elaborated.
On the basis of produced estimation theory the grounds of economic theory can
be generated as well.
Literature:
1. .., .. . .:,
, 1978, .25
2. 9001
3. ..
[] / .. // -
. 2006. 3/3(21). . 58-60.
4. []: . 59203 : G06K 9/00 I..,:
I.. - 20021210028; . 12.12.2002; .
15.11.2005, . 11. - 1.
5. , . .
[] : . . - . : 05.13.07 / . . ; [
] . ., 2007. 36 .
6. [Link]
