123
1 INTRODUCTION
In nature, there is an intrinsic hidden harmony
inherent in it which is reflected in the procedures of
intellectual support (IS) of learning process and
onboard intellectual systems (IP) in the form of
mathematical laws and logical structures. It is this
harmony that makes it possible to explain many
physical phenomena and to predict the vessel’s
behavior in emergency situations using combinations
of observations, measurements, and mathematical
analysis.
The development of scientific understanding of the
problems of the evolutionary dynamics of a ship in
emergency situations, that a university graduate of
the navigational faculty has to study, requires a deep
understanding of the modeling and functional
analysis of the interpretation of the ship’s dynamics in
the behavior and control fields of modern catastrophe
theory, computer mathematics and artificial
intelligence (AI) [1 ] - [25]. The basis of this process is
the idealization of the ship’s dynamics in an
emergency situation based on the integrated
management of the learning process methods and
models of which provide a mathematical description
of dynamic processes within the framework of set-
theoretic interpretation of complex systems (Figure 1).
Learning control unit
Expert module
Trainee modul
Situation model
Interface
Cognitive model
Error
diagnostic
unit
Situation unit
Assessment unit
Simulation unit
Trainee
Figure 1. Intellectual support of navigators training process
by means of emergency control methods
Intellectual Technologies in the Field of Fundamental
E
ducation of Navigators
V. Bondarev, V. Volkogon, Y. Nechaev & P. Kovalishin
Baltic Fishing Fleet State Academy
, Kaliningrad, Russia
ABSTRACT: The problem of “reshaping” the fundamental education of navigators in the conditions of
intensive development of modern computer mathematics, intelligent technologies and high-performance
computing is considered. The main attention is paid to the formation of the information-educational
environment that provides intellectual support for the trainee. Examples of the use of intelligent technologies
that contribute to the organization of the learning process as a creative process of building knowledge are
presented.
http://www.transnav.eu
the
International Journal
on Marine Navigation
and Safety of Sea Transportation
Volume 14
Number 1
March 2020
DOI:
10.12716/1001.14.01.14
124
Homeostasis [11] is a functional state of the
emergency control system in which the maintenance
of dynamic constancy is maintained within the
permissible limits of vital functions and parameters of
the system during various changes in the external
environment. The ability of the operator to provide
the homeostasis with the quality of thinking in a
formalized knowledge space determines the human
potential at a specified time interval and the impact of
human potential on the current state of the vessel as
an object of control (OC) - the human factor.
The report discusses some of the most important
aspects of the problem of improving the training of
specialists using the IS system [12], [13], [15].
2 METHODS OF ARTIFICIAL INTELLIGENCE IN
THE SYSTEM OF TRAINING NAVIGATORS
The developed concept of modeling the behavior of
the vessel in the trainee’s IS system determines the
application of the theoretical principles of
interpretation of evolutionary dynamics in complex
systems. On the basis of conceptual decisions a
continuous process of control over the development
of an emergency situation is formalized - integration,
abstraction and generalization- that determine the
strategy of making management decisions.
The conceptual model of processing the flow of
information based on a multi-model complex (MMC)
of vessel's interaction interpretation in a complex
dynamic environment is:
( )
(
) ( ) ( ){ } ( )
RYWDSVMRRFFU
rn
→××=
τ
:,)(
, (1)
where U(F) – is the functional interaction space;
F(R
n
,R
r
) – function, that determines the spaces of
behavior and control in the process of the situation
development; М(τ) – set of models for interpreting
the situation; V(S) – set of elements of the
visualization system; D(W) – set containing
procedures for generating solutions and generating
control actions; Y(R) – set that forms the rules of a
fuzzy formal system (FFS) of the learning process
management; τ∈[t
0,tk] – time interval for
implementation.
Algorithmic relation feedback in the conceptual
model (1) in the IS system is used to model the
formation of control actions in the process of
performing the proposed task. An important aspect of
new ideas, concepts and theoretical principles of the
learning model is universality. The implementation of
the problem of universality consists in the use of a
dynamic theory of catastrophes [10]. Within the
framework of such an interpretation the construction
of a ship’s behavior model is carried out by studying
the facts and phenomena of the interaction dynamics,
transition to situations generalization and planning
actions that provide modeling and software
implementation of computational urgent computing
technology (UC) [23]. The ideas of universality consist
in the study of various interpretations of emergency
situations, the scenarios of which, determine the
actual processes of functioning in the IS system.
Another aspect of universality is that the developed
models open up the possibility of an effective
combination of the achievements of intelligent
technologies and high-performance information
processing tools that indicate a close relationship
between the methodology that exists between the
theory of learning and methods for studying states
and phase transitions in complex systems.
Formation of conceptual solutions during the
operation of the IS system in the framework of
engineering systematology knowledge envisages the
inclusion of two interrelated phases of research of the
task: generating alternative concepts and choosing the
preferred alternative [10], [11].
The main direction of research in the IS system of
the learning process is the construction and
interpretation of MMC with direct and inverse mutual
relations as well as various environmental factors.
Such formalization provides the possibility of solving
direct (forecast) and inverse (control) tasks based on
various models, and greatly simplify the process of
designing multiscale models from separate blocks:
environmental models, behavioral models, interaction
models.
Interpretation of the interaction dynamics is
carried out within the configuration of the computing
complex presented in the form of graph-interpretation
[11]:
( ) ( ) ( )( )
,,,, CEAUEVFG
R
=
(2)
formalizing events (V (Е, U), describing actions in the
IS system, and conditions A (E, C) - in the form of a
logical description of the state of the vessel (Fig. 2).
INTERPRETING SYSTEM
Current situations
Actions in the system based
on the generated control
transformations.
Implementation - using the
principles of multi-mode
control system
Operating conditions
A logical description of the
state of the system based on
a formalized knowledge
system. Implementation -
within the framework of the
principle of competition and
analysis of alternatives
Figure 2. Evolutionary dynamics of IS interpreted system
As follows from this view, the control module of
the software package of the IS system provides
processing of the information flow associated with the
states of the interaction system and its description
conditions in a complex dynamic environment. The
decision-making support (DMS) in the IS system is
implemented within time interval [t
0,tk] which is
matched by a sequence of discrete events.
(
)
[
]
,,
k
ttS
tS
0
∈
(3)
formalized in the framework of the principle of
competition [9], [15].
As a quantitative measure of the “human factor”
the probability is taken that the decision maker (DM)
transforms the emergency situation into a normal
(regular) situation by searching for effective solutions
125
for the operation of the IS system. An effective
measure to reduce the "human factor" on the basis of
modern IS technologies are intelligent learning
systems (ILS) and simulators [15] with a virtual
reality system. Figure 3 shows the sequence of
operations in the operation of an intelligent simulator
based on the onboard intelligent system of a new
generation.
Figure 3. The flow of information during onboard IS
operation as an intelligent simulator
These technologies illustrate the modern approach
to the organization of IS methods reducing the role of
the "human factor" in making management decisions,
the support of which, is provided by the software
package and the system of controlled remote
experiment [2].
Thus, with the help of the considered intellectual
technologies in the trainee’s IS system, the operator’s
control activity is modeled to prevent and eliminate
emergency situations through the interaction of
natural and artificial intelligence in the DMS system
[1], [2]. The control actions developed by the DMS
system allow the interaction process to be carried out
by managing various types of resources (material,
energy, informational, psychological).
The development of theoretical principles for the
interpretation of complex processes and phenomena
in modern research and practical applications is
characterized by the integration of intelligent
technologies of the 21st century and high-
performance computing. Such integration provides a
solution to complex scientific and technical problems
in the face of uncertainty and incompleteness of the
initial information. The methodological aspects of
developing the IS system [15] are implemented on the
basis of conceptual solutions of the dynamic theory of
catastrophes [10]. This theory opens up the
possibilities of functional analysis and mathematical
logic in analytic and geometric interpretation of the
physical aspects of various applications in the spaces
of behavior and control, defined by the processes of
identification, approximation and prediction. Most
applications are based on the integration of
knowledge about the behavior of the studied
phenomena on the basis of conceptual models,
ontological principles and axiomatics of set theory.
The core of the IS system that supports the creative
process of forming a specialist is a learning model (Fig.
4).
a)
b)
COMPUTING SERVICES
А –external environment
В –vessel dynamics
CONTROL COMPONENT
Computing services management
Management of information
processing
Trainee
С – interaction dynamics
D – modeling of dynamic situations
й
Е – visualization of dynamic scenes
F – generation and analysis of alternatives
G – development of practical recommendations
Visualization of
modelling results
Received data storage
H – risk assessment of decisions
Figure 4. IS model: a - writing of intelligent modules; b -
computing services
The training model is based on information
processing methods using complex ontology, data
mining and revealing “hidden” knowledge (Data
Mining and Image Mining concepts [4]), alternatives
analysis methods and decision generation using Soft
Computing concepts [25], algorithms information
retrieval in structured environments, realized with the
help of domain ontology representation, neural
network technologies, multi-agent, symbolic and
cognitive modeling [9], [11]. The description of the
modules of the trainee IS presented in Fig.4a provides
a solution to the problems of modeling and
interpreting solutions in various practical
applications. The inclusion of new methods and
algorithms for the interpretation of information in the
form of additional modules of the IE creative learning
process will expand the functionality of ILS and will
not change its overall structure and concept of the use
of intelligent technologies. The transformation of
information in the IP modules is carried out within
the framework of a service-oriented architecture [21].
As it can be seen from Figure 4, the tasks of
information processing are the main content of the
five main areas of the use of intelligent technologies
and methods of modern computer mathematics,
implemented by the IS system of the trainee. The end
result of the operation of the IS system is to improve
the quality and shorten the terms of servicing trainees
with the help of software modules. As a result, the
quality of the services provided by the system is
improved, and the directions for the development of
the generated solutions and the multidimensional
analysis of the data obtained are defined using the
concept of “optimal instructor” [6]. The complex
solution of information processing tasks is a
126
combination of hardware and software of the IS
system for creating a dynamic and flexible
information infrastructure that realizes the
capabilities of ILS new generation.
3 FEATURES OF THE TRAINING MODEL IN THE
FORMATION OF A SPECIALIST IN THE FIELD
OF NAVIGATION
The features of the training model in the formation of
a specialist in the field of navigation will be
considered on the example of the functioning of the
training model that provides the IS of the creative
process. We focus on the practical implementation of
conceptual solutions provided in the process of
trainee’s interaction with the program DMS system.
Multi-criteria optimization. The task of multi-
criteria optimization is one of the difficult tasks when
choosing management decisions. The origins of this
task are connected with the famous task of L. Euler
about the seven Koenigsberg bridges (Fig. 5).
Figure 5. Fragment of the situation defining the L. Euler’s
task
The numbers in this figure indicate the numbers of
bridges that need to be bypassed in the absence of
repeated transitions. The solution to this problem is
associated with enormous computational difficulties,
which are defined as “combinatorial explosion”. The
trainee is offered a solution to this problem as applied
to the selection of the optimal ship control strategy
under specified external conditions based on the
genetic algorithm (GA) [11], Bellman – Zadeh
approach [3] and T. Saati’s hierarchy analysis method
[14] with the UC software package. Using the GA
allows you to reduce the task to the search mapping
( ) ( )
StrOptXX
N
→−
1
(4)
within the framework of the optimization structure:
( )
,,: XJRXJ
ϕ
=→
(5)
where (Х
1 – ХN) are vector functions represented as a
given structure; J - target functionality; f:Y→ R -
criterial function; R is a real number characterizing
the specific outcome of the alternative; j (X) is a
function that implements the mapping Х → Y.
Figure 6. The sequence of actions for solving the problem of
choosing solutions for three alternatives
This condition allows the trainee to solve practical
problems of analyzing alternatives when choosing
preferred solutions for a given formalization of a
fuzzy interaction environment, which is especially
important in controlling the dynamics of a vessel
under conditions of intense icing and divergence of
vessels. As an example of working with the software
system, Figure 6 shows the sequence of actions for
solving the problem of choosing solutions for three
alternatives (the program allows up to 10
alternatives).
Multi-agent simulation of the process of
divergence of vessels. The problem of ships diverging
under conditions of intensive traffic flow (TF) is
considered on the basis of multi-agent modeling (Fig.
7) which allows, in the process of learning, to build
dynamically configurable interaction environments of
intelligent agents (IA) depending on the
characteristics of the TF structure.
As a model of the dynamics of the TF, the trainee
is offered a choice of combinations of MASs
depending on the characteristics of the interacting
vessels, including those with an IA leader and with
many agents operating in an integrated virtual
environment [2].
Figure 7. Modeling and visualization of traffic flow
127
This model of MAS allows you to combine certain
aspects of the behavior of interaction objects and has
the flexibility to account for various modifications of
the TF. The input data for the simulation of the MAS
is information on the models of the behavior of the TF
with the interaction parameters and control objectives
in space and time.
4 CONCLUSION
The implementation of the structural and functional
configuration of the IS software package is carried out
using fractal geometry and entropy analysis within
the framework of the information processing
paradigm in a multiprocessor computing
environment [9], [11].
The problem of "embedded intelligence" in the
synthesis of conceptual solutions and practical
applications of research problems "Intelligent
technologies of the twenty-first century" was
discussed during 2009 - 2015. at International
conferences and congresses, including the Forums for
the Development of Modern Society (Section "Science
and Education") in the USA (Washington, San
Francisco) and the UK (Cambridge, Oxford,
Edinburgh).
Expanding the function of "consciousness" and
modeling behavior is the most important evolutionary
task of the trainee. The great Plato said: “Thoughts
rule the world. A thought devoid of striving and
burning is barren”.
Thus, the purpose of this study is to discuss the
main directions of training of specialists in the field of
navigation on the basis of modern approaches to
controlling the dynamics of complex systems in the
framework of the modern theory of catastrophes,
intelligent technologies and high-performance
computing. Conceptual solutions for the
implementation of these problems are based on the
fundamental results formulated on the basis of the
concept of the minimum length of A.N. Kolmogorov's
description [5] within the framework of the
complexity theory [16], the principle of the bifurcation
control by N.N. Moiseyev [8], the theory of incorrect )
tasks of A.N. Tikhonov [17].
REFERENCES
1. Bondarev V.A., Volkogon V.A., Nechaev Yu.I. Conceptual
basis for controlling marine disasters “Current issues
of design, construction and operation of ships and
constructions ", Sevastopol. 2016, pp.28 - 46.
2. Bondarev V.A., Nechaev Yu.I. Artificial Intelligence in
Emergencies seafaring - St. Petersburg: Art Express,
2017. - 336 p.
3. Bellman R., Zade L. Decision Making in Vague
Conditions. - M .: Mir, 1976. - 46 p.
4. Barsegyan AA, Kupriyanov M.S. Stepanenko V.V.,
Kholod I.I. Methods and models of data analysis: OLAP
and Data Mining. - St. Petersburg. BHV-Petersburg,
2004. – 336 p.
5. Kolmogorov A.N. Information theory and theory of
algorithms. - M .: Science, 1987. - 304 p.
6. Krasovsky, A.A., Naumov, A.I. Analytical theory of self-
organizing control systems with a high level of artificial
intelligence // Izv. RAS. Theory and control systems.
2001. No. 1, pp. 69–75.
7. Cook D., Bass G. Computer Mathematics. - M .: Science.
1990. – 272 p.
8. Moiseev N.N. Selected Works, M. Tairex Co., 2003. - 376
p.
9. Neurocomputers in intellectual technologies of the XXI
century. - M .: Radio engineering, 2011. – 352 p.
10. Nechaev Yu.I. Catastrophe theory: a modern approach
to decision making. - St. Petersburg: Art-Express, 2011. -
392 p.. Petersburg: Art Express, 2015. - 325 p.
12. RF patent №2310237 dated 10.11.2007. Intellectual
learning system / Nechaev Yu.I. and etc.
13. RF Patent No. 2356100 dated May 20, 2009. Method of
rating testing in higher education institution / Nechaev
Yu.I. and etc.
14. Saaty T. Decision Making with Dependencies and
Feedback. - I .: LKI, 2008. - 369 p.
15. Systems of artificial intelligence in intellectual
technologies of the XXI century. - St. Petersburg: Art
Express, 2011. - 376 p.
16. Solodovnikov V.V., Tumarkin V.I. The theory of
complexity and design of control systems. - M .: Science,
1990. - 341 p.
17. Tikhonov A.N., Arsenin V.Ya. Methods for solving
incorrect problems. M .: Nauka, 1986.– 285p.
18. Haken G. Information and self-organization.
Macroscopic approach to complex systems. -M .: Kom
Kniga, 2005.– 419 p.
19. Figueira G., Almada-Lobo B. Hybrid simulation –
optimization methods: A taxonomy and discussion //
Simulation Modeling Practice and Theory. - 2014. - V. 46.
- P. 118–134.
20. Foster I., Zhao Y., Raicu I., Lu S. Cloud Computing and
Grid Computing 360-Degree Compared// eprint arXiv:
0901.0131, 2008 [Electronic resource]:
http://arxiv.org/ftp/arxiv/papers/0901/0901.0131.pdf
21. Lublinsky B. Defining SOA as an architectural style. 9
January 2007. [Electronic resource]:
http://www.ibm.com/
developerworks/architecture/library/ar–soastyle/
22. Szalay A. Extreme data-intensive scientific computing //
Computing in Science & Engineering. 2011. - V. 13. - №.
6. - p. 34-41.
23. Urgent Computing Workshop 2007. Argonne National
Lab, University of Chicago, April 25-26, 2007.
24. Wille R. Restructuring lattice theory / ordered sets /
editor I.Rival. - Dordrecht-Boston, 1982.R. 314 - 339.
25. Zadeh L. Fuzzy logic, neural networks and soft
computing // Software on ASM-1994. Vol.37. №3, рр.77–
84.
