#3 / 2020

The questions of qualitative behavior of polynomial vector fields with invariant lines are considered. In particular, we prove that if a polynomial system has an invariant set $M_s^k$ consisting of $s$ and only $s$ parallel invariant lines with an angular coefficient $k$, then $s\leqslant 3$. Moreover, if a polynomial system has invariant sets $M_{n-1}^{k_1},M_{n-1}^{k_2}$ and $M_s^{k_3}$, $(k_1-k_2)(k_1-k_3)(k_2-k_3)\neq 0,$ then for even (odd) $n$ the number of invariant lines of this system does not exceed $2n+1$ ($2n+2$).

11-15.pdf

(355 Kb)

Technical Sciences

The paper presents an analytical review of the methods and algorithms of artificial intelligence. A brief comparative characteristic of the methods has been formed, which allows taking into account their peculiarities and determining the main areas of their application. Comparison of the features of the human brain and artificial intelligence allows us to distinguish two main types of artificial intelligence: strong and weak. This classification makes it possible to analyze the main directions and areas of application of artificial intelligence methods. The paper also discusses the main trends in the development of artificial intelligence and machine learning.

16-25.pdf

(416 Kb)

The article presents a computer model developed by the authors of a wind power plant with a synchronous generator with permanent magnets on a rotor, which makes it possible to evaluate transient processes in various modes when operating on a changing load.

26-33.pdf

(504 Kb)

The basis for constructing systems operating on the principles of image analysis is made up of specific processing methods and their combinations, whose effectiveness has been proven in solving a variety of applied problems. Given the frequency of change and variability of the conditions under which photography is performed, as well as the constant replenishment of the pool of methods available for use, the problem arises of choosing such methods based on the experience of development and application. It is obvious that the components that solve such a selection problem must be based on intelligent technologies. This follows from the fact that descriptions of methods and their application are often not formalized and cannot be processed by classical methods. To solve the problem of choosing the optimal method or a set of methods, the authors solved the problem of compiling a single model based on ontologies, which allows performing tasks such as search, selection and extraction. The proposed ontology contains knowledge about the applicability of methods and their combinations for individual tasks. The ontology is described in the OWL language and is hosted in the cloud, which ensures its continuous availability.

34-39.pdf

(348 Kb)

The article discusses the basic concepts of steganography. The most common methods for hiding data using steganographic algorithms are described. The problem of stability of steganographic systems is described. A technique for detecting information in images is proposed based on the simultaneous use of two methods: signature and statistical. The role of the signature method is to check for possible empty containers in the image and their fullness. If the response is positive, the image is scanned using the chi-square method. A software implementation of detecting embedded information in graphic data is presented. The analysis of various known steganoanalytical methods was carried out and the percentage of hidden information detection in the image was revealed by each algorithm separately. Comparison of the effectiveness of the methods is presented in the form of histograms obtained on the basis of the experiments. Based on the comparison of steganalysis algorithms, the effectiveness of using the combined technique for detecting data in images is substantiated.

40-45.pdf

(400 Kb)

The article analyzes the current state of management information support and reflects current issues of transition to paperless technology. Special attention is paid to the state policy in the field of information technology development. The specificity of systems for collecting, processing and storing information in the information field of higher education institutions is considered. The dependence of the efficiency of the electronic document management system is established depending on the depth and systematic analysis of the composition and structure of information data flows in the organization. The necessity of implementing an electronic document management system, its advantages, opportunities, and prospects is justified. The model, stages and results of implementation of the electronic document management system in the institution are described.

46-51.pdf

(424 Kb)

The current issues of determining failure points in power cable lines and imprecisions that arise when using it are considered. The authors propose a method of improving accuracy of cable damage location due to consideration of its strands twisting and device for its implementation.

52-58.pdf

(330 Kb)

An analysis had been made of the factors on which the intensification of complex natural processes with catastrophic consequences depended. A model for predicting catastrophic consequences of natural processes using a Bayesian belief network is proposed. The peaks of the Bayesian network were highlighted, an expert assessment of the possible values of the indicators was presented and the Bayesian network was trained on the basis of the obtained expert estimates. The “Investment” factor was proposed as a management impact on the network. Modeling and forecasting of possible scenarios for the development of complex natural processes and their catastrophic consequences were implemented. It is proposed to use Bayesian trust networks when building a decision support system for forecasting and assessing risks from damage in the event of catastrophic consequences resulting from the intensification of hazardous natural processes.

59-66.pdf

(332 Kb)

This article examines the vulnerabilities and security threats of a swarm of drones that work together to complete a certain mission. The coordinated actions of a flock of unmanned aerial vehicles (UAVs) can be used to deliver goods and merchandise, serve as mobile hotspots for broadband wireless access, and support surveillance and security of smart cities. The widespread use of drones is also of interest to attackers who are interested in intercepting swarm control and using it to their advantage. The purpose of this review article is to examine possible attacks by attackers – the direction of the attack, possible examples of applications and devices that can be used to carry out the attacks in question.

67-73.pdf

(333 Kb)

The article reflects the current state of the documentation support of technological processes management of educational institutions, and considers topical issues of transition to paperless technology. The necessity of implementing the digital integrated service “Admission to higher education online”, its advantages, opportunities, prospects, and systems for collecting information about issued documents by educational institutions and accumulating this information in the Federal database is justified. Special attention is paid to the state policy in the field of digital technologies development. The article describes modern solutions of all stages of admission to higher education institutions.

74-79.pdf

(549 Kb)

The paper describes the architectural principles of creating a distributed database for the task of perimeter security with high requirements for scalability and fault tolerance. A diagram of the modernized architecture is shown, approaches to achieving the specified key indicators are described, and an approach to migration is developed. A study of the system performance was carried out, and the values of key performance indicators achieved by the fault tolerance subsystem were determined.

80-86.pdf

(575 Kb)

The paper presents recommendations on the use of information (remote) technologies in the process of specialists training in the hospitality industry in the context of the Covid-19 pandemic in the southern region of the Russian Federation. The capabilities of the Moodle 3+ platform to train hospitality professionals are explored. In the context of remote work in the hospitality industry, particular attention is paid to the productivity of specialists. The possi-bilities of using intelligent services on the Moodle 3+ platform are discussed.

87-91.pdf

(298 Kb)

This article discusses the possibilities of using blockchain technology to increase the efficiency of machine learning methods. An analysis of trends in the practical application of neural networks and blockchain technology was also carried out. One of the main problems in the development of neural networks is the formation of training samples. Features of data processing by blockchain platforms provide an opportunity to solve this problem. A practical basis for constructing training samples of neural networks can be information used in blockchain chains. The specificity of data collection by distributed systems avoids a number of typical difficulties encountered in the creation of training samples. Features of the combined technologies impose some restrictions on possible integration methods. Neverthe-less, the discrepancies between the architectures of blockchain series and neural networks can be resolved by dividing the final solution into separate semantic parts.

92-98.pdf

(445 Kb)