t
f
t
H
t
t
t
T
n
n
,
0
t
,
N
n
. (2)
где
1
0
1
2
2
1
2
1
2
1
n
k
t
k
t
k
t
k
n
t
t
i
t
H
,
0
t
,
,..
2
,
1
n
,
t
e
i
k
t
k
,
t
ie
i
k
t
k
t
k
t
k
2
sin
1
1
,
n
k
2
,
0
,
n
.
Из-за вышеуказанных рассуждений следует, что уравнение (2) однозначно разрешима
для любых
N
n
и
0
2
L
f
и решением уравнений (2) является функция
k
k
n
k
k
n
t
f
c
t
1
*
Теорема 1.
Если
Z
, то для
любого
N
n
система линейных алгебраических уравнений
)
(
1
2
)
(
1
2
)
(
1
2
)
(
1
2
)
(
1
)
(
1
)
(
1
)
(
1
)
(
0
)
(
0
)
(
0
)
(
0
.....
..........
..........
..........
..........
t
n
t
n
n
t
n
t
n
t
t
n
t
t
t
t
n
t
t
f
H
f
H
f
H
(3)
однозначно разрешима для любого
0
2
L
f
относительно
,...,
,
1
0
t
t
t
n
1
2
..,
.
решение
t
n
n
t
0
*
*
системы (3) сходиться к решению
t
*
уравнению (3) в норме
пространства
0
2
L
, при этом имеет место оценка
2
1
*
*
;
;
2
0
2
L
f
E
Z
n
L
n
,
где
Z
;
- расстояние от
до множества
Z
.
Наряду с этим приминение этого метода к гиперсингулярным интегральным уравнениям с ядрами
Коши и Гилберта
первого рода
приведены соответственно в работах [3], [4] и [5] а приминение аналогичных
методов к сингулярным интегральным уравнениям с ядрами Коши и Гилберта
приведены соответственно в
работах [1] и [2].
Отметим что, этот метод имеют многочисленные применения в других
областях математики и
механики
Литература
1.Алиев Р.А. Новый конструктивный метод решения сингулярных интегральных уравнений // Мат. заметки,
79:2
(2006), 803–
827.
2. Алиев Р.А., Амрахова А.Ф. Конструктивный метод решения сингулярных интегральных уравнений c ядром Гильберта //
Труды Института Математики и Механики Уро РАН,
18:4
(2012), 14–25.
USE OF NEURAL NETWORKS IN THE OIL INDUSTRY
Nahid Guluzada
1
Imran Bayramov
2
1
Azerbaijan State Oil And Industry University
nahid_98@list.ru
2
Azerbaijan State Oil And Industry University
imranb1963@mail.ru
Abstract.
Oil/gas
exploration, drilling, production, and reservoir management are challenging these days since most oil and
gas conventional sources are already discovered and have been producing for many years. That is why petroleum engineers are trying
to use advanced tools such as artificial neural networks (ANNs) to help to make the decision to reduce non-productive time and cost.
A good number of papers about the applications of ANNs in the petroleum literature were reviewed and summarized in tables. This
paper will provide a review of applications of ANNs in petroleum engineering as well as a clear methodology on how to apply the
ANNs for any petroleum application.
Key words:
artificial neural network
,
neural network model
, oil refining, oil and gas industry
The oil and gas industry is one of the high-tech fields based on modern achievements of science and technique.
This has helped to increase the level of automation of the technological process of oil refining and the development of
the management system.
The process management system plays an important role in the production of oil refining, which is a complex
industrial system. There is a great need to ensure stable and error-free
management of the system, safe and regular
operation. Because even one day's standstill in heavy industry means a huge loss of profits for oil refinery.
Neural networks are also widely used in the oil industry. For example, forecasting of oil extraction from a well.
Experimental data on the coking process were processed to obtain low-sulfur coke with the help of neural
networks. A mixture of heavy oil residues was used as the raw material for the process: heavy gas oil of catalytic cracking,
pyrolysis and heavy tar. The different ratios of these components to the raw material mixture
and the pressure in the
reactor are variable parameters of this process. After processing the data, a neural network that adequately described the
coking process was obtained. Comparing the results of the two methods, it was found that the computational data are
more closely distributed with the experimental values used in neural networks.
Neural networks can be considered as modern computing systems that change information about the image of
the processes taking place in the human brain. The processed data is numerical. This allows to use as an object model
with completely independent properties from neural network. With the help of special programs for modeling neural
networks, it is possible to test it in theory.
Do'stlaringiz bilan baham: