Application of predicate logic for failure detection in SCADA systems
We consider the task of failure detection and localization. It is based on the analysis of the information flow state change in the system. We suggest a structural and logical model to describe SCADA of any topology. It is possible to form diagnostic features of independent failure detection. They a...
Gespeichert in:
| Datum: | 2017 |
|---|---|
| Hauptverfasser: | , , |
| Format: | Artikel |
| Sprache: | English |
| Veröffentlicht: |
Інститут проблем штучного інтелекту МОН України та НАН України
2017
|
| Schriftenreihe: | Штучний інтелект |
| Schlagworte: | |
| Online Zugang: | http://dspace.nbuv.gov.ua/handle/123456789/162349 |
| Tags: |
Tag hinzufügen
Keine Tags, Fügen Sie den ersten Tag hinzu!
|
| Назва журналу: | Digital Library of Periodicals of National Academy of Sciences of Ukraine |
| Zitieren: | Application of predicate logic for failure detection in SCADA systems / M. Alekseyev, I. Udovyk, O. Syrotkina // Штучний інтелект. — 2017. — № 3-4. — С. 150-157. — Бібліогр.: 5 назв. — англ. |
Institution
Digital Library of Periodicals of National Academy of Sciences of Ukraine| id |
irk-123456789-162349 |
|---|---|
| record_format |
dspace |
| spelling |
irk-123456789-1623492020-01-08T01:25:37Z Application of predicate logic for failure detection in SCADA systems Alekseyev, M. Udovyk, I. Syrotkina, O. Прикладні інтелектуальні технології та системи We consider the task of failure detection and localization. It is based on the analysis of the information flow state change in the system. We suggest a structural and logical model to describe SCADA of any topology. It is possible to form diagnostic features of independent failure detection. They are based on the characteristic functions of three-valued logic. We determine the predicate system of knowledge representation to implement the method of SCADA diagnostics in the event of incomplete data. Розглядається задача виявлення та локалізації відмов у SCADA на основі аналізу зміни стану інформаційних потоків у системі. Пропонується структурно-логічна модель опису SCADA будь-якої топології. На основі характеристичних функцій тризначної логіки формуються діагностичні ознаки виявлення незалежної відмови. Визначається предикатна система подання знань для реалізації методу діагностики працездатності SCADA в умовах неповних даних / недостовірних даних. 2017 Article Application of predicate logic for failure detection in SCADA systems / M. Alekseyev, I. Udovyk, O. Syrotkina // Штучний інтелект. — 2017. — № 3-4. — С. 150-157. — Бібліогр.: 5 назв. — англ. 1561-5359 http://dspace.nbuv.gov.ua/handle/123456789/162349 681.518.5 en Штучний інтелект Інститут проблем штучного інтелекту МОН України та НАН України |
| institution |
Digital Library of Periodicals of National Academy of Sciences of Ukraine |
| collection |
DSpace DC |
| language |
English |
| topic |
Прикладні інтелектуальні технології та системи Прикладні інтелектуальні технології та системи |
| spellingShingle |
Прикладні інтелектуальні технології та системи Прикладні інтелектуальні технології та системи Alekseyev, M. Udovyk, I. Syrotkina, O. Application of predicate logic for failure detection in SCADA systems Штучний інтелект |
| description |
We consider the task of failure detection and localization. It is based on the analysis of the information flow state change in the system. We suggest a structural and logical model to describe SCADA of any topology. It is possible to form diagnostic features of independent failure detection. They are based on the characteristic functions of three-valued logic. We determine the predicate system of knowledge representation to implement the method of SCADA diagnostics in the event of incomplete data. |
| format |
Article |
| author |
Alekseyev, M. Udovyk, I. Syrotkina, O. |
| author_facet |
Alekseyev, M. Udovyk, I. Syrotkina, O. |
| author_sort |
Alekseyev, M. |
| title |
Application of predicate logic for failure detection in SCADA systems |
| title_short |
Application of predicate logic for failure detection in SCADA systems |
| title_full |
Application of predicate logic for failure detection in SCADA systems |
| title_fullStr |
Application of predicate logic for failure detection in SCADA systems |
| title_full_unstemmed |
Application of predicate logic for failure detection in SCADA systems |
| title_sort |
application of predicate logic for failure detection in scada systems |
| publisher |
Інститут проблем штучного інтелекту МОН України та НАН України |
| publishDate |
2017 |
| topic_facet |
Прикладні інтелектуальні технології та системи |
| url |
http://dspace.nbuv.gov.ua/handle/123456789/162349 |
| citation_txt |
Application of predicate logic for failure detection in SCADA systems / M. Alekseyev, I. Udovyk, O. Syrotkina // Штучний інтелект. — 2017. — № 3-4. — С. 150-157. — Бібліогр.: 5 назв. — англ. |
| series |
Штучний інтелект |
| work_keys_str_mv |
AT alekseyevm applicationofpredicatelogicforfailuredetectioninscadasystems AT udovyki applicationofpredicatelogicforfailuredetectioninscadasystems AT syrotkinao applicationofpredicatelogicforfailuredetectioninscadasystems |
| first_indexed |
2025-07-14T14:54:46Z |
| last_indexed |
2025-07-14T14:54:46Z |
| _version_ |
1837634564139253760 |
| fulltext |
ISSN 1561-5359. Штучний інтелект, 2017, № 2
150 © M. Alekseyev, I. Udovyk, O. Syrotkina
UDC 681.518.5
M. Alekseyev, I. Udovyk, O. Syrotkina
State Higher Educational Institution “National Mining University”, Ukraine
19, Dmytra Yavornytshogo av., Dnipro, 49027
APPLICATION OF PREDICATE LOGIC FOR FAILURE
DETECTION IN SCADA SYSTEMS
М. Алексєєв, І. Удовик, О. Сироткіна
Державний вищий навчальний заклад «Національний гірничий університет», Україна
пр. Дмитра Яворницького 19, Дніпро, 49027
ЗАСТОСУВАННЯ ЛОГІКИ ПРЕДИКАТІВ ДЛЯ ВИЯВЛЕННЯ
ВІДМОВ У SCADA СИСТЕМАХ
We consider the task of failure detection and localization. It is based on the analysis of the information
flow state change in the system. We suggest a structural and logical model to describe SCADA of any
topology. It is possible to form diagnostic features of independent failure detection. They are based on the
characteristic functions of three-valued logic. We determine the predicate system of knowledge
representation to implement the method of SCADA diagnostics in the event of incomplete data.
Keywords: predicate system of knowledge representation, structural and logical model, three-valued
logic, independent failure.
Розглядається задача виявлення та локалізації відмов у SCADA на основі аналізу зміни стану
інформаційних потоків у системі. Пропонується структурно-логічна модель опису SCADA будь-якої
топології. На основі характеристичних функцій тризначної логіки формуються діагностичні ознаки
виявлення незалежної відмови. Визначається предикатна система подання знань для реалізації методу
діагностики працездатності SCADA в умовах неповних даних / недостовірних даних.
Ключові слова: предикатна система подання знань, структурно-логічна модель, тризначна
логіка, незалежна відмова.
Introduction
Considering the application of expert systems to diagnose SCADA performance, it
should be noted that the relevant task is the development of a reliable and fast decision
support system which significantly depends on the chosen method of knowledge
representation [1–3].
All knowledge representation systems can be divided into the following main
classes: declarative, procedural and special. Predicative systems refer to declarative
knowledge representation systems. It is possible to distinguish procedures to find solutions
(known as a generation mechanism) and procedures to optimize this search (management
mechanism) for declarative knowledge representation systems.
Declarative systems are characterized by the universality of knowledge
representation. The control mechanism, which determines the semantics of the declarative
system and heuristic efficiency to search the solution, reduces the universality of
knowledge representation. Thus, there is a contradiction between universality and
efficiency of knowledge representation for declarative systems [1–3].
Publication analysis regarding topic research
We analyzed the latest research in the field of SCADA diagnostics using expert
system methodology. It showed that today’s expert diagnostic systems are focused on
Technological Control Object (TCO) diagnostics. At the same time, they do not diagnose
the whole SCADA system. Vast, intensive flows of low-level diagnostic information
generated by SCADA causes significant difficulties in its processing by operational
ISSN 1561-5359. Штучний інтелект, 2017, № 2
© M. Alekseyev, I. Udovyk, O. Syrotkina 151
personnel. Therefore, there is a need to implement expert systems as decision support
systems for SCADA diagnostics in real time.
Problem statement
The pressing problem is automatic high-level SCADA diagnostics based on the
methodology of expert systems in real time.
The aim of the research is to increase the quality of SCADA functioning by
developing a method of automatic failure detection and localization in real time. It is based
on the analysis of information flow change when passing through SCADA structural
elements and hierarchy levels. To do this it is necessary to develop a knowledge
representation system which can universally describe the following elements: SCADA
structure of any topology; distribution of diagnostic features for independent failure
detection through structural elements of different hierarchy levels; effective diagnosis
search in real time.
Main part
Consider an example of a given fragment of SCADA structure (see Fig. 1) [4].
Fig. 1. SCADA structural scheme
The set of controlled parameters (CP TCO) at a point of time t is as follows:
)(),(,),(),()( )(21 txtxtxtxtX Xni . (1)
Controlled parameters are measured by sensors and are transmitted to RTUs through
data transmission channels (Ch1). These controlled parameters are then transmitted to
servers through Ch2.
We can apply the following non-decreasing sequences of natural numbers to describe the
distribution of controlled parameters through structural elements of different hierarchy levels [5].
The sequence Kx determines the distribution of controlled parameters through sensors
and Ch1:
Nmx kkkkK ,,,,, 21 . (2)
The sequence Ix determines the distribution of controlled parameters through RTUs
and Ch2:
Njx iiiiI ,,,,, 21 . (3)
The sequence Mk determines the distribution of sensors and Ch1 through RTUs and Ch2:
Njk mmmmM ,,,,, 21 . (4)
ISSN 1561-5359. Штучний інтелект, 2017, № 2
152 © M. Alekseyev, I. Udovyk, O. Syrotkina
We define certain predicates of connection between structural elements of different
hierarchy levels using formulas (2) – (4).
The predicate of connection between CP TCO xi and Sensorµ:
1 1( , ) : (( ) & ( ))?(( 1)?(( ) &( )) : ( )) : 0
Nm NH i i k m i k i k i k , (5)
where mj ϵ Kx.
The predicate of connection between CP TCO xi and RTUj:
0:))(:))(&)?(()1?(())(&)((:),( 12 jjjN iiiiiijNjiijiH , (6)
where ij ϵ Ix.
The predicate of connection between Sensori and RTUj:
0:))(:))(&)?(()1?(())(&)((:),( 13 jjjN mimimijNjmijiH
, (7)
where mj ϵ Mk.
We developed a method of automatic failure detection and localization in SCADA. The
input data in this method are diagnostic matrix D(t). The matrix is represented as a dump
containing the diagnostic features of information flows. It is formed with the sample rate of
data from sensors. The number of rows in this matrix corresponds to the number of SCADA
hierarchy levels. The number of columns corresponds to the number of controlled parameters.
)(1
)(1,1)(
2,1,0,)(
)]([)(
11
33,
,
XniC
SlllSliL
EEtd
tdtD
iCil
iCiL
, (8)
where iL – the index of the matrix row D(t) which corresponds to SCADA hierarchy levels
l; iC – the index of the matrix column D(t) which corresponds to the index of the
controlled parameter xiC(t); l(S1) – the hierarchy level of servers; n(X) – the number of
controlled parameters.
The controlled parameter can have one of three states at each SCADA hierarchy
level: “Absent,” “Non-reliable,” “Reliable.” These states can be described by using Post’s
three-valued logic.
We apply the elementary function of three-valued logic φe – the characteristic
function of the first kind with value e to analyze diagnostic matrix D(t).
3 3
3 3
1, , , {0,1, 2}
( )
0, , , {0,1, 2}
e
x e e E E
x
x e e E E
. (9)
We define diagnostic features for failure detection as follows:
a) A sufficient diagnostic feature of failure absence for the SCADA structural
element (iL, iC) at a point of time t is:
1))(( ,2 td iCiL ; (10)
b) A necessary but insufficient diagnostic feature of failure detection for the SCADA
structural element (iL, iC) at a point of time t is:
1))(( ,2 td iCiL ; (11)
c) A necessary but insufficient diagnostic feature of failure detection due to the
absence of controlled parameters at a hierarchy level (this corresponds to backbone nodes)
or due to absence of data transmission process (this corresponds to data transmission
channels Ch1/Ch2) for the SCADA structural element (iL, iC) at a point of time t is:
ISSN 1561-5359. Штучний інтелект, 2017, № 2
© M. Alekseyev, I. Udovyk, O. Syrotkina 153
1))(( ,0 td iCiL ; (12)
d) A necessary but insufficient diagnostic feature of failure detection due to the
unreliability of controlled parameters at a hierarchy level (this corresponds to backbone
nodes) or due to the unreliability of data transmission (this corresponds to data
transmission channels Ch1/Ch2) for the SCADA structural element (iL, iC) at a point of
time t is:
1))(( ,1 td iCiL . (13)
Analyzing diagnostic matrix D(t) we can assert that no failures have been detected at
a point of time t if the following expression is true for the first row (iL = 1) of diagnostic
matrix D(t) which corresponds to the server’s hierarchy level l(S1):
1))((& ,12
1
td iC
i
iC
N
. (14)
In general, the function of failure detection based on the analysis of diagnostic matrix
D(t) is as follows:
)))((&(),,,( ,22 tdtiLg iCiL
iC
, (15)
where iL – the index of the matrix row D(t) which corresponds to SCADA hierarchy levels
l; α, β – the initial and final ordinal numbers of controlled parameters which pass through
the system’s structural elements for the given hierarchy level l.
Consider the predicate S(i,y,l) to form the criteria of diagnostic feature distribution
through independent failures taking into account the characteristic attributes for each
SCADA hierarchy level. This predicate determines the state y for the controlled parameter
xi at the hierarchy level l:
)(:),,( ,1)1( ilSldylyiS . (16)
Then the diagnostic feature of failure detection can be described by the predicate
S(i,y,l) as follows:
a) ),2,( liS ; b) ),2,( liS ; c) ),0,( liS ; d) ),1,( liS .
It should be noted that both diagnostic features of failure detection ¬φ2(diL,iC(t)) and
the function of failure detection in SCADA g2(iL,α,β,t) do not distinguish independent and
secondary failures. We assume that all the failures are independent at the lowest level lmin
for the given controlled parameter when passing through SCADA hierarchy levels. Thus,
all diagnostic features of failure detection refer to these features of independent failure
detection at level l (l < lmin). We also assume that diagnostic features at hierarchy levels
which correspond to data transmission channels Ch1/Ch2 are diagnostic features of
independent failures.
Therefore, at this stage of diagnostic matrix D(t) analysis we can assert the following:
– The absence of diagnostic features for failure detection at a certain SCADA hierarchy
level is a sufficient condition that no failures have been detected at this hierarchy level;
– The presence of diagnostic features at hierarchy level lmin is a sufficient condition that
there are independent failures at hierarchy level lmin and all the diagnostic features of
failure detection refer to independent failures;
– The number of independent failures at a hierarchy level of sensors (lmin = 2) is equal to
the number of diagnostic features for failure detection;
ISSN 1561-5359. Штучний інтелект, 2017, № 2
154 © M. Alekseyev, I. Udovyk, O. Syrotkina
– In order to define the number of failures at a certain low hierarchy level (lmin > 2), the
additional analysis of diagnostic matrix D(t) is necessary because various diagnostic
features can refer to the same failure;
– The presence of diagnostic features for failure detection at hierarchy level l (l > lmin)
for lL1 is a necessary but insufficient condition of having an independent failure
from low hierarchy levels. It is necessary to have additional diagnostic criteria to
consider it an independent or secondary failure;
– The absence of diagnostic features for independent failure detection at hierarchy level l (l
> lmin) is a sufficient condition of having no independent failures at this hierarchy level;
– To define the number of independent failures when having diagnostic features for
independent failure detection at hierarchy level l (l > lmin), it is necessary to conduct
an additional analysis of diagnostic matrix D(t) because different diagnostic features
can refer to the same failure.
We define the lowest level of SCADA for failure detection lmin in accordance with
SCADA structure (see Fig. 1).
?)),,1,1((: 2min tigiL N
?)),,1,2((( 2 tig N
?)),,1,3((( 2 tig N (17)
?)),,1,4((( 2 tig N
0:)1:)2:)3:)4:5?)),,1,5((( 2 tig N .
If iLmin = 0, then no failures have been detected at a point of time t. Otherwise, the
lowest hierarchy level of failure detection in SCADA is as follows:
min1min 1)( iLSll . (18)
Since for the considered structural and logical model of failure detection and
localization we accept that all the diagnostic features for failure detection ))(( ,2 td iCiL
at the lowest hierarchy level lmin refer to independent failures, then we can form a matrix of
markers with independent failures ʌ(t) for hierarchy level lmin:
))(()( ,2,min
tdt iCiLiCiL . (19)
If lmin < l(S1), then it is necessary to define other SCADA hierarchy levels lmin< l ≤
l(S1). For these hierarchy levels we can detect independent failures when analyzing current
diagnostic matrix D(t). It is possible to take into account permissible changes of the
controlled parameter state when passing up through SCADA hierarchy levels.
Consider the algorithm of independent failure detection in the event of SCADA low
level for failure detection belonging to backbone nodes.
For row iL of diagnostic matrix D(t) the number of diagnostic features ))(( ,0 td iCiL and
the number of diagnostic features ))(( ,1 td iCiL can be defined using the following formulas:
Ni
iC
iCiLN tdtiiLn
1
,0 ))((),,1,(
0
, (20)
Ni
iC
iCiLN tdtiiLn
1
,1 ))((),,1,(
1
. (21)
The total number of diagnostic features for failure detection ))(( ,2 td iCiL for row
ISSN 1561-5359. Штучний інтелект, 2017, № 2
© M. Alekseyev, I. Udovyk, O. Syrotkina 155
iL of diagnostic matrix D(t) is as follows:
),,1,(),,1,(),,1,(
102
tiiLntiiLntiiLn NNN
. (22)
According to the logic of SCADA functioning, at the system’s upper hierarchy levels
lmin< lh+1 which refer to backbone nodes, a necessary but insufficient criterion of having
independent failures is an increase in the number of diagnostic features for failure
detection compared to the system’s lower hierarchy level lmin ≤ lh.
),,1,(),,1,(),,1,(
1)(
,2,1),(2
102
1
1min
tiiLntiiLntiiLn
lSliL
hSlhll
NhNhNh
hh
h
. (23)
If at the system’s upper hierarchy level lh+1, the number of diagnostic features for
failure detection increases compared to the system’s lower hierarchy level lh, we can make
a conclusion that the necessary condition of having independent failures was fulfilled at
SCADA hierarchy level lh+1.
)0),,1,(),,1,(()0),,1,(),,1,(()( 1110101 tiiLntiiLntiiLntiiLniL NhNhNhNhh . (24)
If 1)( 1 hiL , then we can calculate the number of diagnostic features of
independent failure detection at SCADA hierarchy level lh+1L1.
The function of distinction between independent and secondary failures when
controlled parameters pass through SCADA hierarchy levels taking into account the result
of transmitting and receiving data process between adjacent hierarchy levels is as follows:
),3(mod)22222(),,( 22222222222222
4 zyxzxyyzxzyxyxxyzzxyyzxxyzzyxf (25)
where x – the controlled parameter state at a transmitting hierarchy level, y – the result of
transmitting and receiving data process between adjacent hierarchy levels; z – the
controlled parameter state at a receiving hierarchy level; 1),,(4 zyxf – a necessary but
insufficient condition of having diagnostic features for independent failure detection;
0),,(4 zyxf – a sufficient condition of absence of diagnostic features for independent
failure detection.
We can define the number of diagnostic features for independent failure detection at
hierarchy level )( 11min Slll h on the basis of formula 24:
))(),(),((),,1,( ,,1,2
1
44
tdtdtdftiiLn iCiLiCiLiCiL
i
iC
Nf
N
. (26)
If 0),,1,(
4
tiiLn Nf , then this criterion is a sufficient condition that no failures have
been detected at a current hierarchy level.
If 0),,1,(
4
tiiLn Nf , then this criterion is a sufficient condition of having
independent failures.
We can form a row marker matrix of independent failures ʌ(t) for a current hierarchy level:
))(),(),(()( ,,1,24, tdtdtdft iCiLiCiLiCiLiCiL . (27)
Accordingly, predicate ),( liM of having independent failure markers for controlled
parameter xi at hierarchy level l is as follows:
ilSliM ,1)1(:),( . (28)
We can define certain predicates of diagnostic feature distribution through SCADA
ISSN 1561-5359. Штучний інтелект, 2017, № 2
156 © M. Alekseyev, I. Udovyk, O. Syrotkina
structural elements taking into account the characteristic attributes for each hierarchy level.
The predicate of existence of at least one controlled parameter xi at hierarchy level l which
has a diagnostic feature for independent failure detection is as follows:
)),2,(&),((:),(1 liSliMiliiP . (29)
The predicate of existence of at least one controlled parameter xi at hierarchy level l
which has a diagnostic feature for independent failure detection with value y is as follows:
)),,(&),((:),,(11 lyiSliMilyiiP . (30)
The predicate of existence of at least two different controlled parameters xi and xj at
hierarchy level l which have a diagnostic feature for independent failure detection is as follows:
)),(&),(&)((:),,( 112 ljPliPjijiljijPi . (31)
The predicate of existence of at least two controlled parameters xi and xj at hierarchy
level l which have different diagnostic features for failure detection is as follows:
)))),0,(&),1,(()),1,(&)1,0,(((&),(&),(&)((:),,(3 ljSliSljSiSljMliMjijiljijPi . (32)
Thus, for the structural and logical model we consider, the number of independent
failures )32( llnF for controlled parameters with timestamp t at hierarchy level
Sensors/Ch1 can be defined as follows:
– We verify whether there are at least two different controlled parameters xi and xj which
have diagnostic features of independent failure detection at hierarchy level Sensors/Ch1.
Then we verify there are no Sensors/Ch1 for which we have at least two controlled
parameters xi and xj having different diagnostic features for independent failure
detection. This means that the number of independent failures at a current hierarchy
level is equal to the number of hierarchy modules for which we have at least one
diagnostic feature of independent failure detection;
– We verify whether there is at least one Sensors/Ch1 at hierarchy level Sensors/Ch1 for
which there are at least two controlled parameters xi and xj having different diagnostic
features for independent failure detection. This means that the number of independent
failures at a current hierarchy level is calculated by the number of different diagnostic
features for independent failure detection per structural module.
The number of independent failures for the levels of SCADA hierarchy is determined
in analogy to hierarchy level Sensors/Ch1. The foregoing is achieved by taking into
account the connection between various hierarchy levels.
Conclusions
The system of predicates we considered can be applied when forming a knowledge
base of an expert diagnostic system. It allows us to implement a method for SCADA
failure diagnostics. It takes into account the consistencies of information flow changes in
real time in the event of incomplete / unreliable / absent data in the system’s structural
elements. This method of independent failure detection and localization ensures the
reliability of SCADA operational monitoring.
References
1. Giarratano J. Expert Systems: Principles and Programming / J. Giarratano, G. Riley. – [4th Edition]. –
Course Technology, 2004. – P. 842.
2. Varlamov O. Practical Guide on Creation of Miwitary Expert Systems / O. Varlamov, M. Chibirova,
A. Khadiev, P. Antonov, G. Sergushin, I. Shoshev, K. Nazarov. – Tutorial. – M: NII MIVAR, 2016. – P. 184.
3. Ruchkin V. Universal Artificial Intelligence and Expert Systems / V. Ruchkin, V. Fulin. – St. Petersburg:
BHV-Peterburg, 2009. – P. 240.
ISSN 1561-5359. Штучний інтелект, 2017, № 2
© M. Alekseyev, I. Udovyk, O. Syrotkina 157
4. Syrotkina O. Software Diagnostics for Reliability of SCADA Structural Elements / O. Syrotkina,
M. Alekseyev // Power Engineering and Information Technologies in Technical Objects Controls: Taylor
& Francis Group, London. – 2016. – P. 259–265.
5. Syrotkina O. Automatic Diagnosis Method for SCADA Operability / O. Syrotkina // Quality Control
Tools and Techniques. – Ivano-Frankivsk, 2015. – V. 1. – P. 19–26.
РЕЗЮМЕ
М. Алексєєв, І. Удовик, О. Сироткіна
Застосування логіки предикатів для виявлення відмов у SCADA системах
У даній статті розглядається задача виявлення та локалізації відмов у SCADA в
режимі реального часу на основі аналізу зміни стану інформаційних потоків системи
у процесі їх проходження за структурними елементами та рівнями ієрархії. Великий
обсяг та інтенсивний потік низькорівневої діагностичної інформації, що генерується
SCADA, вимагає розробки універсальної та ефективної системи подання знань
стосовно до експертної діагностичної системи підтримки прийняття рішень.
Розглядається розроблена предикатна система подання знань, перевагами якої є
простота реалізації та універсальність опису задачі.
Пропонується структурно-логічна модель для опису SCADA системи будь-якої
топології. В рамках даної моделі визначаються предикати наявності зв'язку між
структурними елементами системи різних рівнів ієрархії.
На основі характеристичних функцій тризначної логіки формуються необхідні
та достатні діагностичні ознаки виявлення / відсутності відмови у системі,
розмежування незалежних і вторинних відмов.
Визначається предикатна система подання знань для реалізації методу
діагностики працездатності SCADA в умовах неповних даних / недостовірних даних.
Ефективний алгоритм пошуку рішення на основі запропонованої системи предикатів
дозволяє проводити оперативний контроль стану структурних елементів SCADA.
Надійшла до редакції 31.10.2017
|