=Paper= {{Paper |id=Vol-2870/paper87 |storemode=property |title=An Option of Building the Distance Learning System with Artificial Intelligence Elements |pdfUrl=https://ceur-ws.org/Vol-2870/paper87.pdf |volume=Vol-2870 |authors=Serhii Lienkov,Sergii Gakhovych,Igor Tolok,Genadiy Zhyrov,Valentyn Bakhvalov |dblpUrl=https://dblp.org/rec/conf/colins/LienkovGTZB21 }} ==An Option of Building the Distance Learning System with Artificial Intelligence Elements== https://ceur-ws.org/Vol-2870/paper87.pdf

An Option of Building the Distance Learning System with
Artificial Intelligence Elements
Serhii Lienkova, Sergii Gakhovycha, Igor Toloka, Genadiy Zhyrovb and Valentyn Bakhvalova
a
Military Institute of Taras Shevchenko National University of Kyiv, Lomonosova str., 81, Kyiv, Ukraine
b
Taras Shevchenko National University of Kyiv, Glushkova ave., 4g, Kyiv, Ukraine

Abstract
The article considers the scientific problem solution of ensuring the required quality training
of the specialists who acquire practical knowledge with the help of computer training
equipment or intelligent computer training systems using distance learning technologies. The
offered decision is directed on the construction distance training system of experts on
intensive training technologies. The offered option of training intensification by the principle
of "inner conviction". This variant of distance learning system construction may provide
preparation of future experts to the maximum possible education level in the shortest time. In
addition, the adaptively controlled algorithm variant to management future experts’ training
process with the best results and the minimum preparation time was proposed.

Keywords 1
Distance Learning System, Artificial Intelligence Elements, Intelligent System, Adaptively
Controlled System, Intensive Training, Educational Task, Computer Training Systems.

1. Introduction
The tempestuous computer technology development and its application in various activities have
led to an inability of teaching future professionals in higher education institutions without computer
training systems. Today developing the computer distance learning system (DLS) and methods of its
use is an important task. Also the DLS may become a foundation for an intelligent training system for
training personnel, which would best meet their purpose. Due to the increasing complexity and
information saturation of different professionals training methods, we meet a need for effective
management of the learning process. As the training system becomes more complex, multifunctional
and is designed for different categories of users, there is a need for adapting it to each user’s
characteristics. The ability of adapting DLS to the user is one of the indicators of its effectiveness and,
as a consequence intelligence. To ensure this user adaptation models of the trainee are developed,
which store information about each user and also the learning management process [1].

2. Related Works
Nowadays a huge number of programs that to some extent increase the effectiveness of learning
through the organization of adaptive dialogue with the user (as a pupil, student, listener, and even
teacher). However, at the present stage, the volume of information growing tempestuously, thus, we
need such e-learning tools that would allow the user not only to view information of interest by
navigating hyperstructures but also to ask various complex questions. It leads to expanding of the user

COLINS-2021: 5th International Conference on Computational Linguistics and Intelligent Systems, April 22–23, 2021, Kharkiv, Ukraine
EMAIL: lenkov_s@ukr.net (S. Lienkov); 4sergah@ukr.net (S. Gakhovych); igortolok@72gmail.com (I. Tolok); genna-g@ukr.net
(G. Zhyrov); val3@ukr.net (V. Bakhvalov)
ORCID: 0000-0001-7689-239Х (S. Lienkov); 0000-0002-9135-6568 (S. Gakhovych); 0000-0001-6309-9608 (I.Tolok); 0000-0001-7648-
7992 (G. Zhyrov); 0000-0001-6388-864X (V. Bakhvalov)
2021 Copyright for this paper by its authors.
Use permitted under Creative Commons License Attribution 4.0 International (CC BY 4.0).
CEUR Workshop Proceedings (CEUR-WS.org)
questions typology, which saves the user’s time in searching information. Thus, in resources [2-3] we
can see general information about LMS systems and a description of all LMS systems’ aspects, such
as their general structure, types, and so on. In particular, the very popular Moodle LMS system was
described in details. There are aspects of the learning styles theory together with the analysis of
students’ behaviour in distance learning described [4]. Another resource [5] propose an adaptive e-
learning system, which able to create learning trajectories adapted to the student’s profile.
Unfortunately, a user has a long time to look through the electronic directory in search of an
answer to one question in many modern systems. The complexity of the nomenclature and the user’s
question require the improvement of describing the educational material options, i.e. except the
structure and content, it is also necessary to take into account the semantic connections between the
described concepts. Using this approach allows to develop of intelligent reference systems or systems
with elements of artificial intelligence, which are needed not only in the DLS but also in any
computer system [6, 7]. Artificial intelligence systems are a type of computer system for effective
training and educating of specialists. The specific feature of such systems is a knowledge base, which
stores the solution to many problems, including the laboratory and practical tasks that the curriculum
integrates. The knowledge base is constantly replenished and modify in accordance with the
experience and requirements of consumers.

3. Methods
The analysis of the existing scientific and methodological apparatus of DLS construction shows its
focus on an extensive path of learning (due to an increase of classes and training number), which has
almost exhausted itself. On the other hand, modern information technologies and teaching methods
with the DLS using open up huge opportunities in the field of intensive educating. As evidenced by
foreign and domestic experience, there is only one rational way: the construction of DLS intensive
training (by improving the quality characteristics of training while minimizing material, resource and
time costs). In this regard, there is a necessity for further development of methodological foundations
for the computer DLS construction with elements of artificial intelligence. This task means the
solution of the following scientific problems:
1. Development of methods for analyzing the process of computer adaptive-controlled DLS
functioning.
2. The development of methodology for structural and parametric computer DLS’s synthesis with
artificial intelligence elements and organization of their optimal functioning in the learning
process.
3. The development of requirements to DLS’s foundation methodical bases on the offered
methodology of their synthesis.
4. The development of DLS’s constructing and functioning methodical principles and their basic
elements that provides realization of requirements for it.
5. The development of an algorithm for specialists’ training process adaptive control.
In general, the task of scientific research formulates as follows. We need to determine a rational
option for the DLS construction, which provide training of future professionals to the highest possible
educating level in the shortest time. The solution to this problem is possible by the use of modern
intensive training technologies. In this case, intensive technology is defined "as a system of factors
that intensify the learning process: ideal, aimed at increasing the activity of trainee, and material
(technical), which providing a given level of training in the shortest time" [1, 8, 9]

4. Results and Discussion
In process of the first scientific problem solving, we define the purpose and model of DLS
functioning and the problems solved by it; conditions of DLS functioning in the future specialists’
intensive training process.
The basic concepts of DLS construction’s methodical bases founded on categories of the general
systems theory [10-13]. At the same time, the commonality of the systems theory concepts is
manifested in a specific subject area, among them intelligent computer simulation systems. The main
elements of DLS consider to be intelligent computer-based learning systems (CBLS), combined with
hierarchical, informational, control links, where each element is aimed at achieving a common goal of
providing the necessary level of future professionals’ knowledge. In the DLS creating process
determining its structure and range of algorithms that best meet the purpose of the system is
necessary. To solve such applied problems knowledge about the effectiveness of DLS structural
organization methods and methods of managing its operation processes are needed. The properties
and patterns of dynamic adaptive functioning, placed in computer networks with different
organizations in the future professionals training, are the subject of the DLS theory. Its main tasks are
the analysis and synthesis of intellectual CBLS.
The tasks of the first group, the analysis of DLS, are characterized by the following two stages of
research. The first stage connects with building a conceptual model of DLS operation, and the second
stage connects with building the dynamic adaptive-controlled process’ mathematical model, which is
created on the basis of the accepted conceptual model. The dynamic adaptive-controlled process
consists of: the object of management – future specialists; controlled parameters – trainee’s activities
quality indicators; control influences – educational tasks (ET), formed by intellectual CBLS; the
algorithm for controlling the learning process – the algorithm for changing the intensity of the ET in
accordance with the level of future professionals training. In addition, the system of indicators and
methods for assessing the effectiveness of intelligent CBLS are also being developed. The analysis
provided within the framework of the DLS construction theory show processes models, its
functioning and regularities which are inherent in these processes or system in general.
The basis of the proposed methodological intensification approach is that the "inner conviction" of
future professionals in the limited time left to study and perform the educational task, causes them a
state of tension. If the tension does not exceed the limit value Vij, the maximum allowable tension, the
impact becomes organizing [14, 15]. In the DLS functioning model, the tension (hij) is defined as the
internal state of the j specialist immediately before the execution of the i elementary task.
The concept of tension is realized in DLS by reducing the cycle of educational information display
(educational task content) on PC monitors until the tension reaches a given level V, at which the lack
of time acts as an organizing factor. The organizing influence of emotional tension (S-tension) is
determined by the fact that in the learning process specialists work more focused, more precisely, and
the probability of correct and timely execution of ET increases.
The tension function (h) is the ratio of the time required to complete the training task to the actual
available time that have specialists in each cycle of DLS
I
 tiTp
h  i 1 (1)
T
where tiTp is the average time required for specialists to perform the i-th ET; І is the number of tasks
remaining to be performed; T is the full time available to future professionals to perform the І
learning tasks remaining in each cycle of DLS functioning [12].
The average values of elementary educational tasks’ performance time are calculated on the
statistical data received during trainings. The calculations of the intensity hij value in the training
process are limited in the DLS by the range from 1.0 to 4.0.
In the general case, the problem of intensive training is as follows: to find from the set of possible
options (X) such a variant of construction DLS (x), which provides intensive future professionals
training (h -> V) to the required (maximum possible) level (P) with minimal financial (C) and time
(T) costs.
The formal statement of the scientific problem is presented in the form of a solution to the
following problem:
P  max, C  min, T  min, h  max (2)
xX xX xX xX

As we see from (2), all the criteria are contradictory and it is extremely difficult to find the best
option for constructing a DLS that satisfies all the above conditions. Without the development of
methodical such systems construction’s bases, structure optimization methods and parameters of
DLS, the decision of the specified problem is impossible.
The conceptual basis for the intensive training DLS construction is a combined solution of the
following two main tasks [16, 17]:
 Accelerated specialists training to the required level of the educational tasks’ implementation
with minimal time costs (the first phase of intensive training);
 Training of specialists to perform tasks up to the maximum possible level of professional
training with given time (cost) constraints during the planned training sessions (the second phase
of intensive training).
The solution to the first problem depend on test results when the specialists have not reached the
required level of training. If it is, the PC of those who are taught are connected to the teacher PC in
order to organize the first phase of intensive training. Thus, the various educational tasks on the
personal computer forms in accelerated mode in such quantity at which the time (financial) expenses
for preparation of the required level expert is reduced.
On the other hand, an adaptive structure of DLS is created for those who have successfully passed
the test. This DLS structure forms the necessary number of educational tasks on the PC display of
trainee and provides further maximum training level increasing.
In the second problems group of the intellectual DLS construction theory, one of the main is the
task of their optimal synthesis, which is aimed at choosing the building a system method that is best
suited to perform the given functions. The initial data in the synthesis problem are next: functions and
tasks of the system; list of restrictions on system characteristics (time, resource); efficiency criterion
that establishes the method of assessing the system quality as a whole. Based on this information, we
can determine the system structure, the parameters of the elements and the processes management
strategy, which must meet the given constraints and be optimal in terms of the content of performance
criteria. The procedure for the synthesis of DLS is divided into procedures of structural and
parametric synthesis. The purpose of structural synthesis is determining the structure of the system:
the subsystems type, the composition of the elements and the relationships between them. The
purpose of parametric synthesis is determining the optimal way of subsystems technical
characteristics and basic elements with a fixed structural scheme of the system. The problem of
synthesis of the optimal structure is considered as the problem of determining the optimal, mapping
the set of performed DLS functions to the set of its interdependent elements.
Thus, implementation of intensive learning technology closely connected with an adaptive change
of the educational tasks speed issuing from the teacher’s PC to the trainee PC. In this case, the
structure of the DLS is conventionally divided into computer (training) classes, which include the
teacher’s PC plus the trainee PC and the distance learning subsystem, which includes the teacher’s PC
and the PC of future professionals. The physical connection of the PC to the DLS may be provided
through the internal computer network intranet or the global computer network Internet, which
provides a flexible intellectual DLS structure for the most efficient conduct of all computer training
types.
The method of solving the parametric synthesis problem includes the following main stages. First,
the efficiency indicators of each options constructing DLS on a set of system functioning conditions
to select the best of them are determined. At the second stage, the problem of situations classification
on the satisfaction basis with the accepted restriction is solved. Under the situation in the
multidimensional factor, space means a solution, as well as the conditions of its implementation. For
each situation’s point, performance indicators are calculated and the obtained values are compared
with the allowable ones. The third stage implies narrowing the set of solution options that is achieved
by using the Pareto optimization principle, which distinguishes the allowable Pareto-effective set of
solutions [16, 17]. Further narrowing of the solutions set is associated with the conceptual choice of
such a DLS construction variant from the whole set, which provides a sufficiently high (necessary)
level of target and economic efficiency. Thus, the optimum decision help to define a range of
intelligent training system parameters admissible values that provides preparation of future experts to
the necessary level on performance of the set tasks.
Methodical bases of requirements substantiation to DLS are based on the methodology of their
synthesis and include the decision of the third problems group: training level dynamics modelling in
the course of future experts’ preparation and according to the received indicators of target and
economic efficiency, so requirements to DLS construction are formulated. Also, the requirements for
DLS architecture, knowledge base, expert subsystems of objective control, planning and management
of training and information environment, hardware and software of all DLS subsystems are
substantiated.
The fourth group of tasks involves the development of DLS construction technical principles, i.e.
technical solutions that will ensure the implementation of reasonable requirements for them. A
conceptual point of view shows DLS is an integrated intellectual-adaptive system that first studies
educational tasks, and then with the help of artificial intelligence elements at the stage of initial
training educates specialists to perform these tasks. As the necessary skills and work abilities are
learnt, the information support of artificial intelligence changes adaptively.
The main purpose of DLS is to train professionals to the required (maximum possible) level (Рн)
with minimal time (Т) and resources (C).
In this case, the generalized indicator C should take into account the costs of development (С1),
serial production (С2) and implementation (С3) of each r-th (r = 1, ..., R) DLS variant, time (С4), as
well as operational costs (С5) required for the optimum level specialists training (Рн). In addition, the
generalized indicator C should take into account the costs of creating and maintaining databases
(knowledge bases) of learning tasks (С6), the organization of objective control and learning process
management (С7). Also, the generalized indicator C may include the costs (С8) required to increase
the stability of the computer software functioning and network equipment of each r-th DLS variant.
In this case, the value of the k-th (k =1, s) cost indicator should not exceed the maximum allowable
value of СКдоп.
The cost indicators are set in different units and have different physical meaning, so for choosing a
rational option for building and organizing the DLS functioning in the first phase of training let us use
the concept of compromise’s nonlinear scheme [16, 18].
In process of selecting the r-th (rational) version of the DLS, which provides accelerated training,
it is advisable to use the following generalized indicator (Сr):
s  F C 
Cr    k kдоп   min (3)
k 1  Ckдоп  Ckr 
s
with Pr > Рн, СКr < СКдоп,  Fk  1 ;
k 1

where Pr is the average level of specialists training, which is achieved by using the r-th DLS version
in the first phase of training; Рн is the required level of specialists training; Fk is the coefficient of the
k-th indicator importance.
In addition, the r-th (rational) DLS version construction, in the second phase of training, must
meet the following performance criteria:
Рr –> max, (4)
s
at СКr < СКдоп,  Fk  1 , (r =1, … , R; k = 1, … s),
k 1

The initial stage of any complex system designing is building its architecture and knowledge base,
which depends on the means and methods of this system’s implementation.
The option for organizing the adaptive management of the specialists training process follows
below.
The goals of the learning process’ adaptive management may be substantiated by the main stages
of organization and conducting specialists training using adaptive training tools (Figure 1).
At the first stage, the goals and tasks of training are formulated. The ultimate goal of such training
is to educate staff to the required level, which ensures the implementation of educational tasks in the
most difficult conditions.
If the staff have reached an "excellent" training level as a result of previous education, then the
main purpose of further training is to maintain the previously achieved level.
If the staff have not reached a "satisfactory" level in previous training, then the purpose of each
subsequent training is to achieve a "satisfactory", "good", and then "excellent" level of training. The
level of trainees training is characterized by its ability to timely and accurately perform standard
operations at a given intensity of training tasks.
The average number of simulated situations per time unit that require personnel’s intervention
characterizes the intensity of the learning tasks flow.
Consider the main stages of organization and specialists’ training with using adaptive training
tools to substantiate the goals of the learning process adaptive management (Fig. 1).

Figure 1: The main stages of organization and specialists training using adaptive DLS

At the first stage, the goals and tasks of training are formulated. The ultimate goal of such training
is to educate staff to the required level, which ensures the implementation of educational tasks in the
most difficult conditions.
If the staff have reached an "excellent" training level as a result of previous education, then the
main purpose of further training is to maintain the previously achieved level.
If the staff have not reached a "satisfactory" level in previous training, then the purpose of each
subsequent training is to achieve a "satisfactory", "good", and then "excellent" level of training. The
level of trainees training is characterized by its ability to timely and accurately perform standard
operations at a given intensity of training tasks.
The average number of simulated situations per time unit that require personnel’s intervention
characterizes the intensity of the learning tasks flow.
It is assumed that during simulation of certain typical situations by personnel, in accordance with
the algorithm of its work, certain types of ET operations are performed.
In order to achieve by trainers a training level that allows timely and error-free do ET operations
in the most difficult conditions it is necessary, based on the actual level of the studied, to determine
for each training dose-progressive intensity of simulated situations (XR) exponentially.
Thus, the ultimate goal of training (Zt) is to achieve by staff the required level of training (РТ) to
perform basic ET operations types at a given intensity of simulation situations (TN):
P  PT
ZT :  .
 R  TR R  1,2,...,N 
In the second stage (as shown in Fig.1) the required predicted levels of staff training to perform
the main ET operations types are determined: Р*, …, Рn*. Since any ET contains a large number of
similar operations, it is convenient to use the probability of timely and error-free (correct) execution
of operations Pi*, i = 1, … n as an indicator of the required predicted staff’s knowledge level in the
implementation of the i-th type of operations.
The required level of staff training to perform ET in general (РT) may be described as a function:
РT = РT (P1*,…, Pi*,…, Pn*).
It follows that the required level of staff knowledge to perform ET (РT) depends on the levels of
staff training in the realization of each operation type (Pi*, i = 1, …, n).
The solving of determining the indicators Pi*,…, Pn* on the basis of the established requirements
to the level of staff training for the implementation of the entire ET (РT) task is ambiguous. The
number of all possible solutions (n - 1) is a parametric set.
For example, consider ET, which consists of six operations. If ET must be performed with a
reliability of not less than 0,8, then
РT (P1*,…, Pi*,…, P8*) = 0,8.
This equality is a mathematical record of the requirements and constraints imposed on the level of
staff training to perform eight types of operations. Moreover, the probabilities of seven types of
operations can be chosen arbitrarily, calculating the reliability of only the operation of the i-th type:
Pi* = РT-1 (P1*,…, Pi-1*, Pi+1*,…, P8*). (5)
Since Pi* is the root of equation (5), the number of unknown variables is a seven-parameter
set [19, 20].
It follows that the task of determining the requirements for the levels of staff training to perform
operations of each type has plenty of solutions.
Each of these solutions has its advantages and disadvantages. Firstly, the level of staff knowledge
to perform ET depends on the levels of staff training in the realization of each operation type and,
secondly, there is a relationship between the required projected level of staff training (Pi*) and time
of performing each operation type (Т*, i = 1, … n).
In the process of staff training with using DLS, the above dependencies are usually not taken into
account. The basic types of operations are practised to some extent until the staff reaches the required
level of training to perform ET as a whole (РT). In the best case, the experience of the training head
helps to practice one or another type of operation to the level set by the head. At the same time, the
training manager uses an approximate intuitive, far from optimal, solution, which, in turn, increases
the time of staff training to the required РT level.
The disadvantages of such staff training process management are:
 Subjectivity of the training head both at the organization and at management of training
process;
 The complexity of defining, memorizing and analyzing quantitative staff training levels
indicators to perform basic types of ET operations;
 The difficulty of recording and analyzing the individual staff abilities to perform certain ET
operations types;
 The difficulty of determining the required predicted staff training levels to perform each
operation type without special mathematical apparatus using.
This does not take into account the fact that staff may perform some types of operations in a
timely and error-free manner from the beginning to the end of training. These types of operations are
actually practised throughout the training, although the time of their training, and, consequently, the
total training time could be reduced.
Thus, in the second stage of the training organization and conduct, taking into account the
requirement to perform all ET (РT), from the whole set of possible solutions for each type of
operation is determined such that the training time to the required predicted levels P1*,…, Pn* will be
minimal.
To prepare staff to the above-mentioned knowledge levels, and, consequently, to the required
training level for the implementation of the entire ET (РT), the estimated training time is calculated by
the formula:
n
T =  Ti
i 1

where Тi (і = 1, … n) is the average time of staff training to the required predicted level Pi*.
With
Тi = Si  i
where Si is the minimum required number of the i-th type situations, the simulation of which provide
the i-th type ET operation’s testing and the staff achieve the required predicted level Pi*;  is the
average time of the i-th type situation simulation.
The calculation of Pi*,…, Pn* and Ti*,…, Tn* should be carried out taking into account the
patterns of increasing the staff training level (P) from the time of ET operations each type’s practice.
At the third stage of the organization and providing staff training a situation in accordance to
design, in advance the optimum plan of intensive preparation is simulated.
In order to make the simulated situation as close as possible to the real one, it is necessary to use a
set of such situations that took place in the staff work or may arise in conflict situations.
Direct testing of ET operations main types is carried out in the fourth stage, the essence of which
is as follows.
In the process of simulating a circumstance, the staff performs a finite number of operations types
n. At the same time, depending on the level of staff training to perform standard operations, a certain
type of situations that require intervention in the staff management process is created. In other words,
a simulated circumstance is a source of information for staff, using which they assess the situation,
make decisions and perform appropriate operations types of the training task.
In the process of standard operations realization by staff, an adaptive change of circumstances is
taken place, i.e. adjusting the number (intensity) of simulated situations depending on the staff
training level to perform standard operations (arrow 1 in Fig.1).
At the same time, there is an objective control over the staff work. The assessment of the staff
training level is based on the analysis of probabilistic characteristics, which take into account the
accuracy and timing of the operations main types and ET in general.
For the above-mentioned reason, at certain intervals, the registration of the actual staff training
level to perform each i-th type of operation Pi, (і = 1, … n) and the entire ET (Рбр) are provided:
Р = Р (Р1, Р2, … , Р1, … , Рi, … , Рn).
Simultaneously, a comparison of the actual and required training levels (P and РT) is conducted. If
the condition
Р ≥ РT (6)
performed, the training may be completed as the staff has reached the required knowledge level to
perform ET. In other cases, the following condition is checked:
|Pi* - Pi| ≤ Pi , (і = 1, … n) (7)
where Pi is the value of the maximum allowable error in the implementation of operation ET i-th
type by staff.
If condition (7) is met, it is assumed that the actual indicators of staff training levels Pi, …, Pn with
the established accuracy degree (Pi, і = 1, … n) correspond to the required predicted levels Pi*, … ,
Pn*. Therefore, the practice of typical ET operations continues according to pre-set time intervals
Ti,…, Tn.
If condition (7) is not met, then there is a need for a gradual adjustment of the organizing and
conducting staff training process, i.e. in the adaptive management of the training process.
The realization of such management is needed due to the actual level of staff training or
incorrectly formulated training purposes (ZТ), or insufficient accuracy of the time indicators T1, …, Tn
are determined during which staff are trained to the required predicted levels P1*, … , Pn*. The
explaining of this is primarily the uncertainty (at the initial stages of training organization and
conducting) in the initial data on the staff training individual level in performing the main types of
operations and ET in general. Therefore, at the fifth stage of the training organization and carrying
out, it is necessary to correct (adapt) parameters: of the simulated circumstance S1, ..., Sn; T1, …, Tn of
the required predicted levels Pi*, … , Pn*, as well as РT due to changes in the actual level of staff
training (P).
In other words, as soon as conditions (6) and (7) are not met in the training process, it is necessary
to adjust the indicators P1*, …, Pn* and, respectively, the minimum required number of situations S1,
..., Sn. This simple adaptive control cycle of situation simulating process (arrow 2 in Fig.1) is the
lower adaptation level of the organizing and conducting training process. At this level, an adaptive
change of the simulated circumstance is provided depending on the level of staff training to perform
each ET operations type.
The main purpose of the circumstance’s simulating’s process’s adaptive management is to
minimize the time of staff training for the required predictable levels to perform the main ET
operations types. Thus, adaptive control of circumstance simulation process consists in change of the
simulated situations parameters S1, ..., Sn and, accordingly, working off operation each type for the
minimum time intervals Ti, …, Tn.
If the lower limit of process adaptation (training conduct and organization) is not effective enough,
i.e. conditions (6) and (7) are not constantly met, it is advisable, based on the actual level of staff
training, to adjust training objectives (ZT) and then the corresponding indicators P1*, …, Pn*, T1, …,
Tn (S1, ..., Sn). This cycle of the circumstance’s simulating’s adaptive process control belongs to the
upper level of training (its organization and conduct) processes adaptation (arrow 3 in Fig.1). The
adjustment of the required staff training level (РT), the intensity of situations reproduction (tr) based
on the actual level of staff training to perform ET are upper adaptation level characteristic features.
The main purpose of situation simulation process adaptive control (Zv) is to reduce staff training
time to the required level for the implementation of ET in general:


 P( P1 ,...,Pn )  PT ( P1*,...,Pn * );

Z T :  R   TR R  1,2,...,N  ; (8)
 n
T   Ti  min .

 i 1
According to mentioned conditions (8), an important task of circumstance’s simulating’s process's
adaptive control's implementing the is determining the required predicted levels P1*, …, Pn*.
This problem solving requires the investigation of the dependence between the staff training level
and the time of the ET operations main types and on this basis to develop a method of optimal
requirements distribution for personnel training levels to perform typical ET operations.

5. Conclusions
Thus, in the article are proposed ways to solve the scientific problem of building and organizing
the DLS functioning with artificial intelligence elements for training specialists with using simulators
or intelligent computer training systems for practical and laboratory classes. The requirements to the
intellectual DLS are substantiated by the structural and parametric synthesis problem’s solving and
organization of their optimum functioning in the training course with using the offered criteria. The
algorithm for the adaptive management organization of specialists’ training process using the
predicted training levels in order to reduce training time was offered for training staff with using of
adaptive training tools.

6. References
[1] K. F. Boriak, Yu. O. Hunchenko, S. V. Lenkov, V. Ie. Lukin, S. A. Shvorov, Automated learning
systems with elements of artificial intelligence for the study of technical disciplines, VMV,
Odesa, 2012.
[2] N. Cavus, Distance Learning and Learning Management Systems, Procedia – Social and
Behavioral Sciences 191 (2015) 872–877. doi.org/10.1016/j.sbspro.2015.04.611.
[3] P. Ifinedo, J. Pyke, A. Anwar, Business undergraduates perceived use outcomes of Moodle in a
blended learning environment: The roles of usability factors and external support, Telematics and
Informatics 35(1) (2018) 93–102. doi.org/10.1016/j.tele.2017.10.001.
[4] R. D. Costa, G. F. Souza, R. A.M.Valentim, T. B.Castro, The theory of learning styles applied to
distance learning, Cognitive Systems Research 64 (2020) 134–145. doi.org/10.1016/
j.cogsys.2020.08.004.
[5] M. Boussakssou, B. Hssina, M. Erittali, Towards an Adaptive E-learning System Based on Q-
Learning Algorithm, Procedia Computer Science 170 (2020) 1198-1203.doi.org/10.1016/
j.procs.2020.03.028.
[6] T. H. Teng, A. H. Tan, L. N. Teow, Adaptive computer-generated forces for simulator-based
training, Expert Systems with Applications 40(18) (2013) 7341–7353. doi.org/10.1016/
j.eswa.2013.07.004.
[7] D. Treceño-Fernández, J. Calabia-del-Campo, M. L. Bote-Lorenzo, E. Gómez-Sánchez, R. de
Luis-García, C. Alberola-López, Integration of an intelligent tutoring system in a magnetic
resonance simulator for education: Technical feasibility and user experience, Computer Methods
and Programs in Biomedicine 195 (2020) 105634. doi.org/10.1016/j.cmpb.2020.105634.
[8] N. D. Kriukova, The role and place of the conceptual and terminological apparatus in the
development of the theory of intensive vocational training technology, in: A. P. Beljaeva (Ed.)
Metodologicheskie osnovy proektirovanija intensivnyh tehnologij professional'nogo obuchenija
[Methodological foundations for the design of intensive vocational training technologies], Saint-
Petersburg, 1992, pp. 26–32.
[9] B. M. Herasymov, Design and application of expert-educational systems, European univ. publ.,
Кyiv, 2008.
[10] K. O. Soroka, Fundamentals of systems theory and systems analysis, KhNАМG, Khmelnytskyi
2004.
[11] V. I. Korniienko, O. Yu. Husiev, O. V. Herasina, V. P. Shchokin, Theory of control systems,
NHU, Dnipro, 2017.
[12] Yu. O. Hunchenko, S. V. Lienkov, S. A. Shvorov, A. A. Honcharuk, A method of managing the
process of training special forces specialists at a shooting range, Zbirnyk Kharkivskoho
universytetu Povitrianykh Syl [Bulletin of Ivan Kozhedub Kharkiv Air Force University], 4(33)
(2012) 258–261.
[13] G. Zhyrov, S. Lienkov, Y. Husak, H. Banzak, I. Tolok, Analysis of problem optimization of
parameters maintenance process according to state with constant periodicity of control,
International Journal of Emerging Trends in Engineering Research 8(6) 2020 2606–2611.
doi.org/10.30534/ijeter/2020/63862020.
[14] G. P. Shibanov, Quantitative assessment of a person in human-technology systems,
Mashinostroenie, Мoscow, 1983.
[15] V. Ie. Lukin, Application of automated systems for educational purposes in the laboratory
workshop, in: Abstracts of reports of the All-Ukrainian scientific-practical conference, 20 April
2012, KPU, Zaporizhzhia, 2012, p. 204.
[16] A. M. Voronin, Yu. K. Ziatdinov, A. V. Kharchenko, Complex technical and ergotechnical
systems, Fact, Kharkiv, 1997.
[17] V. Ie. Lukin, S. A. Shvorov, Construction of intelligent learning systems, Zbirnyk naukovykh
prats Viiskovoho instytutu Kyivskoho natsionalnoho universytetu imeni Tarasa Shevchenka
[Bulletin of Military Institure of Taras Shevchenko Kyiv National University] 32 (2011) 143–
147.
[18] A. M. Voronin, Yu. K. Ziatdinov, Vector optimization of dynamic, Tekhnika, Кyiv, 1999.
[19] I. V. Stetsenko, Systems modeling, ChDTU, Cherkasy, 2010.
[20] T. V. Savchenko, S. V. Gakhovych, Theoretical and practical aspects of using the distance
learning system, Zbirnyk naukovykh prats Viiskovoho instytutu Kyivskoho natsionalnoho
universytetu imeni Tarasa Shevchenka [Bulletin of Military Institure of Taras Shevchenko Kyiv
National University] 55 (2017) 186–193.