Table of contents

13th MULTICONFERENCE ON CONTROL PROBLEMS (13 MCCP-2020)

6–8 October 2020

The 13th Multiconference on Control Problems, organized by the State Research Center of the Russian Federation - Concern CSRI Elektropribor (St. Petersburg, Russia) was held on October 6–8, 2020.

Concern CSRI Elektropribor has held the multiconferences together with some research institutions and universities since 2006. The idea of holding the multiconference consists in uniting several local conferences, intimately related by the topics of scientific problems being considered. The Saint Petersburg Group of Russian National Committee on Automatic Control has become the initiator of such cooperation. The content of the Multiconference slightly changes from year to year, but the Conference in Memory of the Prominent Designer of Gyroscopic Devices N.N. Ostryakov, which has been held by CSRI Elektropribor every two years since 1956, remains a permanent part of the program. Traditionally, every even-numbered year the Multiconference is held at Concern CSRI Elektropribor, JSC (St. Petersburg), and every odd-numbered year — in Divnomorskoe town (Gelendzhik, Russia), where the A.V. Kalayev Research Institute of Multi-processor Computation Systems, Southern Federal University, is the main organizer.

On the following page you will find the declaration form.

• Please answer each question.

• You should submit the form along with the rest of your submission files.

• The deadline is the submission date written in your publishing agreement.

All conference organisers/editors are required to declare details about their peer review.

We will published the information you provide as part of your proceedings.

Peer review declaration

All papers published in this volume of Journal of Physics: Conference Series have been peer reviewed through processes administered by the Editors. Reviews were conducted by expert referees to the professional and scientific standards expected of a proceedings journal published by IOP Publishing.

Our reviewers' corpus includes eminent experts, prominent scientists and researchers. We usually use single-blind type of PR, but also in controversial cases we can use double-blind. We send the paper to the reviewer for 3 weeks to make a review, then everything depends on the results. If there is any doubt about acceptance or rejection, we ask another reviewer to analyse. Then we take the responsibility for making a decision.

To facilitate submission we developed an editorial system based on the comsep.ru platform: comsep.ru/conf, where authors, editors and reviewers can easily track the history and progress in reviewing, editing and translating an article. Loading paper is now clearer, simpler and more convenient than ever.

• Type of peer review: Single-blind / Double-blind / Triple-blind / Open / Other (please describe)

• Conference submission management system: comsep.ru/conf

• Number of submissions received: 415

• Number of submissions sent for review: 415

• Number of submissions accepted: 158

• Acceptance Rate (Number of Submissions Accepted / Number of Submissions Received X 100): 38

• Average number of reviews per paper: 3

• Total number of reviewers involved: 54

• Any additional info on review process:

• Contact person for queries:

Name : Dmitriy Taranovskiy

Affiliation: PhD, head of editorial group

Email : editor@eprib.ru (Concern CSRI Elektropribor, JSC, editor's office)

An essential requirement for achieving a high level of performance, autonomy, and reliability of a multi-robot system performing joint long-term operations in an unstructured environment is an advanced control system. In general, the robot's control system is designed to be hierarchical and consists of several subsystems. To unify the interaction of individual components and reduce the load on computing and communication devices, we use the event-based methodology at different levels of the control system. In the paper, we demonstrate how this methodology can be applied to solve four challenging problems in robotics: cooperative formation control, path planning, missions scheduling, and action planning.

Artificial Intelligence, part of Information Technology, is currently the hottest spot in Information technology the main research subject of which is knowledge and the main research directions are methodologies, methods, and algorithms for knowledge acquisition, knowledge representations, and knowledge usage. The paper describes the core scientific content of these directions, outlines their state-of-the-arts, and analyzes briefly new trends and prospects for research and development in Artificial Intelligence for the forthcoming years.

The term "linear quadratic problem" refers to the problems of dynamics where linear controlled systems are associated with a certain quadratic functional of performance. "Linear" approaches are the approaches that are usually applied to linear systems, e.g. the methods of the integral transforms (Laplace, Fourier), the linear functional analysis and the frequency domain methods. The paper considers some cases where these approaches are applied to nonlinear systems.

The main problems of trajectory processing of radar data are considered. Within the framework of the task of multitarget traсking, approaches are proposed that effectively solve the following problems: assignment problems; detection (reset) of trajectories; trajectory measurement of coordinates and motion parameters of the observed objects; highlighting trajectory features.

The concept of an intelligent hybrid aircraft fault-tolerant flight control system operating under complete parametric and structural uncertainty is considered. Fault tolerance in the proposed system is ensured through the efficient integration of model-based and model-free health monitoring and reconfiguring methods.

The article discusses the possibilities of using electric discharge technologies to create sources of acoustic waves in sonar countermeasure systems. Presented are the results of preliminary research works aimed at determining the parameters of acoustic waves generated by electric discharge technologies. The possibility of using electric discharge as a source of acoustic interference for undersea weapon guidance systems is shown.The current situation in the world, the armed conflicts and terrorist threats require ensuring the safety of maritime mobile objects (vessels, ships) from being hit by naval weapons. Today, one of the means of surface or submarine ship destruction is a torpedo weapon. The main means of protection against destruction by torpedo weapons are the ship maneuvering and its sonar countermeasures to the weapon guidance facilities [1]. The methods of sonar countermeasure are divided into passive and active ones [1 – 4]. Active methods are aimed at interfering with guidance systems, interrupting the wake, placing target simulators, etc.

The paper presents the scientific and practical foundations of the developed applied theory of proactive control of information processes in the industrial Internet based on cloud and fog computing. The main scientific contribution of the research results is that modern informatics, thanks to the developed theory, is enriched with the methodology and methodological support developed in classical cybernetics and neocybernetics. The paper shows the main elements of this theory, analyzes the possible ways of its development and practical use.

Various applications of strapdown electrostatic gyroscopes (ESG), depending on the type and purpose of the vehicle on which the gyro is installed, as well as operation conditions, impose substantially different requirements on this gyro, including accuracy of information it provides. One of traditional methods used to ensure the gyro accuracy is self-compensating rotation of the housing. The paper considers some applications of a strapdown ESG with a solid rotor and the self-compensation methods used in these applications. This gyroscope was developed 20 years ago at the CSRI Elektropribor, but the work on its improvement goes on continuously. Theoretical aspects of the ESG operation are discussed, taking into account the developed drift model, as well as their experimental validation based on the analysis of ESG operation in polar orientations.

The development of fiber-optic and electronic components has led to a significant widening of their application scope, in particular in the construction of ground vehicles navigation systems. Often, such systems have the special performance requirements under difficult temperature conditions: the environment temperature changing rate in some cases may reach several °C/min. One of the key elements of a navigation system is a fiber-optic gyroscope (FOG). The harsh environmental conditions increase the demands on production technology of the FOG, on the choice of construction materials, on the arrangement and shape of the parts, etc., however, in order to decrease temperature effects influence on the output signal of the device, the use of special algorithmic compensation methods is also required. Algorithmic compensation requires careful preparation of input data, which is associated with extensive experimental work. This article analyses the FOG temperature dependence characteristics in the wake of the designated experimental preparation for algorithmic compensation.

To adjust the stabilization systems that ensure the operation of the angular position or rate sensor based on the effect of nuclear magnetic resonance, it is required to choose the criteria for the quality assessing of sensor's operation. The article presents the criteria of the quality of the angular sensor operating as a self-oscillation system. It is shown that the accuracy of the angular sensor increases when the proposed criteria are met.

In the presented paper we consider a well-known in-flight alignment problem for strapdown inertial navigation system (INS) via position and velocity data provided by the reference inertial navigation system. Main challenges of the mentioned problem are displacement of the reference navigation system relative to the navigation system being aligned, mutual misalignment of the relevant body frames, the presence of uncontrolled time delays in the positional and velocity aiding data. We propose in-flight alignment algorithm that was tested based on simulation. Typical vehicle's trajectories were derived from relevant telemetry data. The results of numerical experiments are also presented.

The critical analysis is given concerning the current state of using deep neural networks with convolutional and recurrent layers, a recurrent network of Long Short-Term Memory, Gated Recurrent Units for estimation tasks in relation to navigation and motion control. A comparison of neural network and traditional methods is given for understanding and explaining their functioning. The differences, advantages and disadvantages of deep neural networks in relation to solving estimation problems are revealed. The possibility of machine training with reinforcement is analyzed for estimation tasks in navigation and motion control in real time. The prospects of using neural networks in the processing of navigation data, as well as for the tasks of adaptive estimation and trajectory tracking, are formulated.

The paper describes a control system for an autonomous driving car. Control systems are based on deep neural networks. Due to the impossibility of preliminary formation of a high-quality training sample for problems, the use of reinforcement learning is considered. Possible variants of the modeling environment implementation are described.

The current paper demonstrates the effective capabilities of deep neural networks in solving the problem of identification of constellations from a photo of the sky in conditions of a priori uncertainty, incomplete observability and stochastic disturbances. The quality of solution 0,927 by metric F1 is obtained. In order to achieve the result, the original ResNet-like architecture of the convolution neural network was synthesized; statistical analysis of the dataset was carried out, the function of losses and strategy of neural network training were developed, and an accurate criterion of constellation observability in the image was formed. The observation of noise influence on the quality and stability of solutions was carried out.

In the present paper a comparative analysis of two architectural neural network approaches (based on 3D convolutional and LSTM) in the recognition of actions on video is made. The problem was being solved on 10 behavior classes separated from the UCF50 dataset. The original neural network architectures were developed and pre-trained. It was found that the network based on 3D convolutions has better generalization ability and is more stable in the training.

In the work, the functions of cognition are brought in line with the aspects of the functioning of cognitive transport systems and technologies. We consider the mathematical apparatus of cognitive technologies – neural network control systems, based on trained multi-layer neural networks of direct action in the class of adaptive control systems for dynamic objects. The possibility and variants of their inclusion in the control system of the transport system are clarified. On the basis of the considered approach, a country road recognition system based on a convolutional neural network has been developed, which will automatically segment rural roads using satellite images and build a road network diagram to analyze the spatial development of transport networks and to control unmanned vehicles. The development of a neural network based on U-net was carried out in Python 3x. The training set consists of 880 images prepared by hand mark up. The accuracy of the developed model when testing on prepared samples was 64%. According to the results of the study, conclusions were drawn and prospects for further functional development of the developed tools were determined.

The possibility of state estimation for a dynamic system with nonlinearities in the equations of dynamics and measurements using the Singular Transformations Method is discussed. The features of the proposed approach application are considered. For a methodological example, the equations for the required estimates are specified and the estimation accuracy analysis is carried out.

The paper studies how the phase inductance of brushless DC torque motors affects their performance within high precision electric direct drives. Modified control laws for torque motors with high phase inductance are described, which optimize their performance by various criteria. The modified laws are based on continuous regulation of mutual attitude of current and voltage vectors in the motor windings. Parameters of torque motors are compared with the described control laws applied, and advantages of their use are covered.

The paper analyses the torque characteristics, angular errors, natural frequencies, and finite elasticities of electromechatronic module within gearless servo system with coaxially arranged multi-pole brushless DC torque motor and high-precision induction resolver. We provide analytical expressions and describe conditions for optimization with regard to the following criteria: sufficient torque, minimized angular errors, downsized construction, and absence of resonant frequencies as applied to principal and design solutions of the module. The results from calculation studies for module implementations with different types of torque motors are discussed.

A new dynamic model of a distributed real-time computing system is investigated. The model is embedded in the system software and is intended for test diagnostics of violations in addressing exchanges between the system software modules. The questions of observability and controllability of the proposed model are discussed for the case when models of links are used without loss of information.

An approach to scheduling a computational process in distributed real-time systems with jitter minimization is considered. The approach is based on the concept of a solvable class of systems for which there exist optimal scheduling algorithms of polynomial complexity.

The article is devoted to the idea of combining navigation and measurement software to provide a higher level of efficiency and quality of aerial electromagnetic survey. Complex geophysical measurement systems of "Impulse-Aero" series are overviewed. Current survey technology with the above-mentioned systems includes human-based control of the measurements with a geophysicist on-board. The technology also includes the navigational support by the geophysicist or an additional navigation specialist. The concept of automatic combined complex for navigational and measurement monitoring is proposed in the article. The main goal of the system under analysis is to automatically estimate the quality of received data and to give necessary recommendations on the screen of the navigation unit. As the result of the survey, some tasks during the flight can be solved by pilots without any additional staff. This approach is supposed to increase flight safety, reduce costs and raise the efficiency of the survey in general.

Technological solutions based onthe integrated approach are presented, which optimize the manufacturing process and satisfy the requirements imposed on spherical rotors of electrostatic gyros. System modelingis used to study the main stages of defining the rotor functional parameters and to constructa system of models that determines the coordinated processes of diffusion welding, precision sphere lapping and rotor weight correction. The practical results of rotor manufacturing and the prospects for accuracy enhancing are provided.

To estimate the derivatives of a deterministic signal arriving in real time, a state observer is introduced, built as a replica of the virtual canonical model with an unknown bounded input. The results of a comparative analysis of observers with linear and piecewise linear corrective actions are presented, which are confirmed by the results of numerical modeling for piecewise differentiable composite function. Its derivatives have jump discontinuity points at the moments when the form of the function changes. It is shown that at these moments the observer variables with linear corrective actions with high gains have large surges, which increase by an order of magnitude with an increase in the order of the estimated derivative. Observers with bounded piecewise linear corrective actions do not have these problems and are recommended for use in practical applications.

The spectral method to solve estimation problems for linear continuous-time stochastic systems with polynomial measurements is presented. It is based on both the spectral form of mathematical description (the representation of deterministic functions and random processes by orthogonal series) and the particle filter. The main goal of this work is to implement the continuous-time particle filter without a time discretization. The proposed spectral method provides the possibility to solve estimation problems such as filtering, smoothing and prediction.

Observers of disturbances and measurement noises are developed for dynamical systems with sector-bound nonlinearity and parametric uncertainty. Conditions for the existence of these observers are proposed in the form of solvability of linear matrix inequalities (LMIs). Using the observers, a control law is designed to ensure system stabilization. The proposed algorithm is experimentally studied on a stand for multi-machine power systems control.

The paper presents a new adaptive observer of the local coordinates of a moving object based on measurements of linear velocity, yaw angle (heading), and distance to a single beacon with known position. The conditions under which the algorithm is asymptotically stable are shown, as well as the states for which a robust estimate is reasonable to be used.

Comparative analysis of two algorithms for underwater vehicle positioning using measurements of range to a drifting buoy and radial velocity is carried out. The algorithms based on building location contours and polynomial filtering are considered. Invalid data rate and radial RMS position error are analyzed to compare the algorithms. Based on the analysis made, the polynomial filtering algorithm is accepted as preferable.

An adaptive magnetic flux and position observer has been designed for a permanent magnet synchronous motor. The measurable signals are stator winding currents and voltages only. The observer provides observation error convergence to zero in finite time for measurements without noise. The design method is iteratively applied to obtain the rotor velocity and load torque estimates.

In this paper, we study the problem of designing a linear controller for a stationary discrete system with a small step. By the method of integral manifolds, the discrete optimal control problem with a small step is subdivided into two subproblems, the solutions of which are found independently of each other. Algorithms for analytical solutions of subproblems are based on the second Lyapunov method. As a result, a linear controller was designed for a controlled stationary discrete system with a small step, consisting of two subcontrollers, which separately regulate the movements of a system of lower dimension. The results obtained can be used to design an optimal linear digital controller.

The paper proposes a method for the synthesis of nonlinear systems invariant to the output variable, under the conditions of parametric and external perturbations, taking into account the constraints on phase variables and control. To solve the problem, the ideology of the block method using sat-functions as local feedback, and the disturbances observer on the sliding modes are applied.

In the present paper authors consider the synthesis of combined nonlinear control law for non-affine control plant with delay in the input variable with step-by-step changing of its dynamics. The control plant operates in conditions of a priori parametric and structural uncertainties in the presence of external noise and measurement of the controlled variable only. The structure of the control system includes the predictor-compensator and the filter-corrector.

This work presents new ideas related to the synthesis of nonlinear stabilizing control laws in the framework of optimization approach. The focus is done on the optimal damping concept, proposed by V.I. Zubov in the early 60s of the last century. This theory is used to reduce significant computational costs in solving optimal stabilization problems. The essential features of the aforementioned concept are taken into account, allowing the construction of new methods for practical synthesis of control systems with desired dynamic properties. Various modern aspects of the optimal damping theory's practical implementation are discussed. Special attention is paid to the specific choice of the functional to be damped to provide the desirable stability and performance features of the closed-loop connection.

This paper is devoted to inverse problems of the control theory, namely, the dynamic control reconstruction problem. It is the problem of online reconstruction of unknown controls (the input) using known inaccurate measurements of the realized trajectory (the output). Deterministic affine controlled systems are considered. A method for solving this problem is suggested. It relies on auxiliary variational problems on extremum of a regularized integral residual functional. The key feature of this method is using a functional which is convex in control variables and concave in state variables. Properties of the solutions obtained by this method are studied. It is shown that the obtained solutions have oscillating character and are bounded. Results of numerical simulations are provided.

A method of synthesis of modal control in servo systems with a differentiating observer of the reference action is proposed. It is shown that with the joint design of the modal controller and the observer, the uncertainty of the synthesis problem arises, which leads to the oscillation of the regulated value under step setting actions. To eliminate this uncertainty, the transition to the optimization problem of control synthesis was carried out. A numerical example of the formulation and solution of this problem is presented, which confirms the effectiveness of the proposed method.

The paper proposes a method for solving the Brockett problem based on the synthesis of robust modal control for systems with non-stationary parameters, which does not require complex adaptation algorithms that involve the introduction of a variable regulator matrix. A transition to the task of synthesizing the regulator is shown, using a certain optimality criterion with limitations, which provides similar values to those of the regulator coefficients. A practical numerical example of setting and solving this problem is presented, which confirms the effectiveness of the proposed method. To illustrate the results obtained, a root travel time curve was constructed for two systems: robust and non-robust.

The problem of ensuring the accuracy for constant and harmonic input actions in systems with modal control is considered. An analytical method of combined formation of both the control accuracy and the type of transient process is proposed. The modal control possibilities for harmonic signals are shown, for which the synthesis method is usually applied to the desired frequency response. It also introduces a requirement of quality in terms of speed. This problem is being solved by drawing up an inequality system. These inequalities are illustrated graphically. Using a numerical example, the effectiveness of the proposed method for solving the problem is shown.

The paper considers the task of the automatic positioning of the mobile robot in front of the external visual marker in the case of the presence of the transport delay in the control channel. As a basic regulator without the delay compensation we use mixed approach based on the visual servoing method and a special multipurpose control structure which is capable of taking into account a set of requirements to the controlled dynamics. To compensate the delay we apply the modification of the basic regulator which keeps the original transfer matrix. We also demonstrate the efficiency of the proposed approach by the results of the experiments with the computer model.

A study of the stability of a multiply connected automatic control system containing digital links by the frequency method is proposed. The stability criterion is based on the system description of the dynamic properties of the system at the level of characteristics of subsystems and characteristics of communication.

Dynamics of the New Keynesian model in continuous time with the Rotemberg pricing mechanism is considered within a framework of an optimal control problem. Various regimes of monetary and fiscal policy ('active' and 'passive') can lead to unstable dynamics in the economy. Parameters of the Taylor rules for both monetary and fiscal policies determine conditions for local equilibrium determinacy. Mapping out the ranges of the Taylor coefficient values where local determinacy cannot be obtained allows to control the economic system by controlling these parameters.

The paper considers the problem of analytical design of optimal controllers in the most general formulation for nonlinear plants with affinely (linearly and additively) input states and/or controls. On the basis of a new approach to approximately optimal control synthesis, a generalized theory of analytical design of optimal controllers is being developed. The main theorems are formulated; new concrete forms of non-classical quality functionals are obtained.

The paper describes a method and tools for prototyping components of intelligent systems. These means are the parts of the PESoT technology based on generative and visual programming as well as the model transformations. The proposed method redefines the main stages and models of the standardized model-driven approach, in particular, Model Driven Architecture in the context of developing knowledge bases and expert systems that use the logical rules formalism. The description of the method is presented. Proposed tools include languages and software. Specifically, the Rule Visual Modeling Language and Personal Knowledge Base Designer are used for knowledge bases engineering including modeling and codification; the Transformation Model Representation Language and Knowledge Base Development System support model transformations and the rapid creation of computation-independent models. Described tools were applied to solve problems in the field of reliability and safety of technical systems.

Execution of large-scale search and survey operations by groups of autonomous mobile robots requires the ability to respond to unforeseen events of different nature in a quick and efficient manner. In order to achieve this requirement, the available information about recent changes should be spread among all active group members as soon as possible. Informing a non-stationary group of mobile robots distributed across different regions of the operational area is a nontrivial task for large groups with limited communication capabilities. The paper addresses the problem of agent-based information distribution throughout a network in a new original formulation. In essence, the problem is to define an effective order of transferring data between single robots in order to achieve the fastest way of sharing it among the whole group. A number of heuristic strategies are proposed for building initial solutions alongside with the local search scheme for their further optimization. The comparison between the suggested approaches is made regarding both the solution quality and the speed of algorithm operation. Some issues related to the test instances generation are also considered.

Continuous improvements in the hardware, software and algorithms of unmanned vehicles (UVs) are nearing their mass deployment in all physical environments, including within unmanned groups, and with traditional means of transport controlled by the operators (drivers). Notwithstanding the measures taken to improve security and resilience, UVs as complicated cyberphysical systems have a number of vulnerabilities and failure points. This paper comprises the review of known vulnerabilities and failure points (including those in group control scenarios), and the development of the proposals to improve the resilience and security of UVs, and addresses the identified issues.

The paper is devoted to comparison of the efficiency of information exchange by autonomous mobile agents when searching in a limited coverage area. The special feature of the problem is to consider the case of operation within the prohibition of usage of radio, light, and acoustic communication channels at a great distance, as well as without surfacing. The paper compares the following algorithms for the work of groups of autonomous robots, based on a preliminary division of area into sectors and survey of robots in their sectors using the "breadth-first search" method.

The mathematical models describing behaviour of an operators network, and in particular multi-stage processes, are quite common in practice. In this work, the authors propose approximate methods for solving the control problem with parameters on the network of operators. They are based on the methodology developed in works [1-4].

This paper investigates a new way to position unmanned aerial vehicles, that act as routers, to allow subscribers to access a communication network. In my model, the subscribers move along known trajectories in sparsely populated areas. To solve this problem, I developed a software to simulate how the subscribers move. I also developed an optimization algorithm that maximizes my custom objective function, where the function sums the probabilities for the subscribers to access the network.

The paper provides a general view on the original logical inference based approach to dealing with discrete event systems as subject to supervisory control theory. The approach suggests a representation of discrete event system as a positively constructed formula to imply automated logical inference in the calculus of positively constructed formulas. During the inference, languages of the discrete event system are generated and analyzed. The minimally restricting supervisor for uncontrollable specification may be also designed. A nonblocking supervisor design is illustrated with a simplified model of autonomous underwater vehicle operational modes switching.

Second-order protocols for consensus and convergence of agent's trajectories to a fixed configuration are investigated. It is proved that the asymptotic behavior of the second-order system is uniquely determined by the eigenprojection of the Laplacian matrix of the communication digraph. The presented protocol and the corresponding expression for the asymptotic behavior are generalizations of the protocols for simpler cases.

A routing problem with precedence conditions and complex cost functions is considered. In this problem, we must choose a starting point, a route (permutation of indices) and a specific trajectory for our process. This trajectory must be consistent with the route. In addition, this route or index permutation defines the sequence of tasks. In addition, the selection of the above route must satisfy the precedence conditions defined by the system of ordered index pairs. These ordered pairs are called address pairs. We consider the additive criterion routing problem. This criterion is natural for the problem of dismantling the system of radiation sources. In this article, we will focus on this engineering problem. In this problem very naturally cost functions arise with a dependency on the list of tasks. Namely, each time the performer touches those and only those sources that were not dismantled at that time. The solution uses widely understood dynamic programming. We build the optimal algorithm for the PC; information about the computational experiment is given.

The article discusses a method for assessing the innovativeness of an enterprise in digital economy, based on the collection of information according to the certain assessment criteria, and also includes a developed mathematical apparatus for calculating the integral index of innovativeness. The article proposes an "Entry-Exit" assessment model.

This paper introduces a constructive definition of an informational community, which agrees with formal models of opinion dynamics for bounded rational agents in social networks. As is shown below, uncertainty can be taken into account when detecting informational communities. An example of a stable informational community is given. A control problem for informational communities is stated, and its solution is presented within the DeGroot model.

The paper is devoted to simulation modelling of the arid plant communities (APC) behaviour. Such systems belong to the class of agroecological systems and include a large number of interacting natural factors that contribute to the system's degradation. We propose a model of APC behaviour based on a weighted oriented graph. The values of weights and other coefficients reflecting the exposure of factors and their mutual relationship are acquired by expert assessments. The proposed model allows predicting the behaviour of APC in response to various biogenic control actions.

A stochastic dynamic system is considered, for which a decision-making system is built over the stabilization loop, an analogue of the second signaling system. For this purpose, the principle of large deviations and the theory of optimal control are used.

In the paper a flight control system for a piloted aircraft on the pitch angle is considered, taking into account the rate saturation of the aircraft controlling surface. To prevent occurrence of the non-linear pilot-involved oscillations (PIO), a nonlinear correcting device is introduced into the control loop. On the basis of the pilot's adaptation to control tasks variations, his/her mathematical model parameters are optimized and the choice of the nonlinear correcting device parameters is proposed. The sensitivity functions of the non-linear system are calculated for harmonic and square-wave reference signals. The results of the optimized system simulations are presented demonstrating good performance of the closed-loop system with nonlinear correction.

A new paradigm in the theory of control systems is presented, in which all the basic concepts and properties of systems are formalized on the basis of a single principle - the general principle of isomorphism. The principle made it possible to formulate and prove the realization theorem, on the basis of which the basic concepts (model, abstract system, simple, complex, large system) have been mathematically strictly defined and the fundamental properties of systems and criteria for their provision have been formalized: in a new way, up to isomorphic models of a rather general form, the concepts and criteria of observability, controllability, checkability, identifiability of systems have been defined, and new formalized methods of analysis and synthesis of regulators, observers and integrated information-control systems have been developed, the examples of applications of the theory have been given.

The branch and bound method is used to obtain exact solutions of the single machine scheduling problem with the objective function of maximum lateness minimization. Lower bounds are estimated by solving dual problems. A machine-independent value, namely the number of branching points in the search tree, is used as the complexity index of instances. The instances are presented as points in 3n-dimensional space. Known and newly developed ways of instances generation (with a large number of branching points) are presented, and the advantages and disadvantages of each of them are described.

The article describes a model of knowledge, called the notional model. A notion is a thought about a certain unique representation from the subject domain. The notional model uses notion-signs, notion-associations and notion-generalizations that are formed by primary mental abstractions: identification, generalization and association, respectively. The notional model consists of the notional structure and the contents of the notions. The notional structure describes each notion as a set of other notions united by one of the primary mental abstractions. The content of the notion is described using enumerating and resolving procedures. The notional model is semantically invariant, improves the transparency of the subject domain, and allows you to create the knowledge base with a linear estimate of a query execution time.

A general approach to decompose the interaction of agents in social, organizational and technical systems is proposed. Within this approach, the problems of reachability, sustainability and searching for an optimal trajectory can be solved independently for each agent.

On the plane, we consider the differential game of two identical pursuers against three consistently evading players (targets) forming a coalition. Players' movements are simple. The speed of the pursuers is limited to one, the speed of the targets is limited by a given constant of less than one. At the beginning of the game, the pursuers and targets are localized at two different points on the plane: the pursuers are at a certain point P⁰, and the targets are at a point E⁰. A game is conducted with complete information. The task of the pursuers is to pinpoint all the targets in the shortest possible time. The task of evading players is the opposite.

n-Person prisoner's dilemma is a Markovian sequence of static games that are realized randomly depending on players' previous choices. The dynamic Shapley value is constructed in this case. The work is a continuation of the paper published earlier by Grinikh A. L. [1]. In this paper, we consider the new approach of dynamics of game which depends on the strategies of all players.

This paper discusses one of the approaches to analysis of the global stability boundaries of type 2 PLLs.

The paper discusses one of the issues of interaction of inverter electrical systems with control systems based on special models of the phase-locked loop, called Enhanced Phase-Locked Loop.

Analysis of the phase space of discontinuous systems via classical methods of the theory of oscillations can be quite a difficult task. Moreover, in some cases this analysis is impossible to perform. Due to the emergence of frequency methods for analysis of discontinuous systems and the development of various analytical-numerical methods, it has become possible to give a more accurate description of the phase space of systems with discontinuous right-hand sides. The present report tries to adopt to the Watt's governor model that takes account of the self-regulation of the object and the derivative action. We compare the results with those obtained by rigorous analytical method.

The paper is devoted to asymptotic behavior of synchronization systems, i.e. Lur'e–type systems with periodic nonlinearities and infinite sets of equilibrum. This class of systems can not be efficiently investigated by standard Lyapunov functions. That is why for synchronization systems several new methods have been elaborated in the framework of Lyapunov direct method. Two of them: the method of periodic Lyapunov functions and the nonlocal reduction method, proved to be rather efficient. In this paper we combine these two methods and the Kalman-Yakubovich-Popov lemma to obtain new frequency–algebraic criteria ensuring Lagrange stability and the convergence of solutions.

Gennady Alekseevich Leonov (1947 – 2018) – Corresponding member of the Russian Academy of Sciences, Foreign Member of the Finnish Academy of Science and Letters, a Council Member of International Federation of Automatic Control, the Highly Cited Mathematician of the Russian Federation, a bearer of many notable awards – passed away on April 23, 2018 after a short battle with a grievous illness. This paper is written in his memory. The authors remember Prof. Leonov as eminent scholar, admired teacher, and unconventional administrator.

A brief review on stabilization problem for linear control systems is presented. The statements of Pyragas' and Brockett's problems for stabilization of linear control systems by time-delayed feedback control and time-varying static output feedback are given. Leonov's algorithms of nonstationary stabilization for solutions of Pyragas' and Brockett's problems are presented. Necessary and sufficient conditions for the stabilizability of two- and three-dimensional systems are given.

In this paper, the anisotropy-based analysis problem for the linear discrete time-invariant system with state-multiplicative noises is considered. The mean anisotropy of stationary input sequence is assumed to be bounded by given nonnegative value. For this setting, the boundedness condition for anisotropic norm of the system is obtained in terms of Riccati equations.

The paper discusses a mathematical model of the second-order measuring transducer, which is based on the Lagrange equation of the second kind. On the basis of the theory of high-order Sobolev-type equations and optimal control of solutions to the Showalter – Sidorov problem, a theorem on the existence of an optimal dynamic measurement is obtained. Second-order sensors such as accelerometers, two-link oscillators and some other systems can be investigated in the framework of this mathematical model.

A semilinear descriptor system with a linear part and a nonlinear term is considered. This system is not resolved with respect to the derivative of the desired vector function. The article describes the optimal measurement problem for the model under study with the Showalter–Sidorov condition. Based on the methods of the theory of dynamic measurements, sufficient conditions for the existence of a solution to the problem of optimal measurement are found.

In this study, one of the main problems of practical implementation of adaptive control and identification systems is considered. This is the problem of choice of an adaptation rate value to provide the required convergence rate of linear regression parameters. The most widely applied method to identify such regression parameter values is the gradient descent (GD) one. But the adaptation rate is usually chosen manually for it and kept constant in the course of the plant functioning. The recursive least-squares (RLS) method can be used instead of GD as it provides the developer with the equation to adjust the adaptation rate automatically, taking into account the current value of the regressor. Therefore, the aim of this research is to compare the known properties of GD and RLS methods and demonstrate the capability of the RLS to provide a constant convergence rate under the condition of the time-varying regressor.

The article contains a description of the measuring device mathematical model by the methods of quasi-normalized spaces. This model serves as an application to the theory of positive degenerate holomorphic groups of operators in Quasi-Sobolev spaces of sequences. This approach is relevant due to the possibility of finding positive solutions or values of the measuring device, which means that it is more consistent with the physical meaning. Indeed, as a rule, we talk about its absence when we get a negative or imaginary value of a quantity that is positive or real in meaning

The theory of optimal dynamic measurements is based on the use of methods of optimal control theory to find a dynamically distorted input signal from the observed output signal. The model of the measuring system contains a Leontief type system, the initial Showalter–Sidorov condition. Inertia of the measuring system and interference lead to distortion of the input signal. We present a numerical method to use a digital moving average filter for the observed signal with subsequent application of the method of optimal dynamic measurement, as well as the results of computational experiments.

The problem of optimal dynamic measurement consists in finding the dynamically distorted input signal by observing the output signal. This problem is solved as an optimal control problem to minimize the difference between the values obtained from the output data and the model values obtained in the course of a computational experiment. It was shown earlier that this problem has a unique solution. For a complex form of the input signal, various numerical methods can be used to find the optimal dynamic measurement. This article suggests to use the spline method in order to solve this problem. Namely, we present the algorithm of this method and show that the solution obtained by the spline method converges to the exact solution of this problem.

A plane problem for optimization of stationary sensors placement in a threat environment to counteract an evading object is considered. For this purpose, an auxiliary task of constructing locally optimal paths of the evading object for a given configuration of sensors is solved. Integral risk functional is minimized for the evading object. The minimum of this functional is required to be maximized in the task of sensors placement. The paper describes the methods of solving both problems and the results of numerical modeling for the case of 2 sensors: examples of local optimal path maps, and the best found configurations of sensors placement.

The principles of application of projection-operator methods of mathematical programming for calculating optimal controls based on "immersion of control problems" in problems of finite-dimensional mathematical programming are formulated. The results obtained illustrate the technique of qualitative study of the stability of systems with optimizing feedback based on the contraction mapping principle of functional analysis. The considered technique allows one to synthesize locally optimal systems with nonlinear control objects. In this case, nonlinear models of objects are taken into account in the right-hand sides of linear manifolds, and the projection operator forms nonlinear optimal controls with feedback taking into account the operators of nonlinearities of the object.

In connection with the wide spreading of various intelligent sensors, IoT devices, smartphones, autonomous transport systems, various industrial and home automation systems, an unprecedented amount of data is generated, including those intelligently linked to each other. Linked data allows you to build complex and varied relationships between objects and subjects of the real world. Unfortunately, modern big data processing systems and machine learning models are extremely poorly suited for working with such dynamically linked data, especially in the case of real-time systems. We discuss current and future-proof approaches to working with such data using graph analysis models.

In the paper, an approach for estimation of objects relative to specified target that was proposed by Wierzbicki A. P. is analyzed. It is asserted that this approach does not contradict with using utility (value) functions and weight coefficients at generalizing function for solving the same task. This assertion is justified by a mutual inverse relationship between functions estimating the goal achievement and deviation from a goal. It is shown that the deviation functions additionally allow solving such problems as ordering objects by penalties and bonuses, as well as by deviation from the norm. Using a simple medical example, the results of ranking patients by deviation from the norm, obtained by the criterion of minimax optimization and the additive generalizing function, are compared with equal and different values of the weighting coefficients of the indicators.

A group behavior and control model in a system consisting of autonomous real-world and Internet objects with opposition is considered. The urgency of the problem is due to the growing role of Internet technologies in competition models in economics, politics, computer security, etc. The problem model is analyzed in the context of the Boyd Cycle. The tasks of the cycle phases are formulated, a multi-agent model of the behavior of autonomous objects and the digital twin of the external environment in the control loop is substantiated. The necessity of using the model of a unified semantic knowledge and data space is argued.

This work is about navigation safety of marine traffic at sea areas. In addition to traditional approach of danger situation detection based on vessels approaching, the current paper introduces another metrics derived from kinematic parameters of the vessel to identify whether it follows patterns (rules) of the traffic at a certain sea area. Authors focused their efforts on analyzing existing traffic schemas in order to identify its danger level in general rather than scrutinizing individual cases. Along with the traditional approach of sea traffic schema identifications, we propose an original method of automated identification of sea traffic schemes based on clustering of movement parameters using historical AIS data. For the clustering decomposition subtraction clustering algorithms were considered. The historical AIS data of sea traffic at Tsugaru strait is used for identifying traffic schema and ship routes planning with the model designed under presented research.

A decision support system (DSS) is described, which helps operators to solve the problem of resource-saving control of extruders of various types used in the production of multi-assortment packaging polymer films (PF). It includes an industrial data mining subsystem for predicting the state of extruders and film quality, a subsystem for evaluating the performance of extruders, an information subsystem, and a data visualization subsystem for building trends of controlled and calculated parameters, interfaces for the extruder operator, administrator and knowledge engineer. The results of testing according to the data of extruders produced by PF at factories in Russia and Germany confirmed the efficiency of the DSS. The use of DSS allows you to save the resource of extrusion equipment, increase the time periods between its stops, and reduce the repair time.

The article discusses possible approaches to data collection and information collection in multi-level distributed systems. It is suggested to use model based approach. The requirements to these models are analyzed and the system of models is proposed. The system of models describes the structure and behavior of the observed system in terms of graphs. The resource graph, the data flow graph, the control flow graph, and the query flow graph are used. These graphs are automatically constructed on the basis of information received from the observed system in the form of log files. A virtual machine that implements this approach is proposed.

This paper considers the problem of control of a multi-degree-of-freedom (multi-DOF) nonlinear electromechanical plant with elastic properties and an external unknown disturbance. The unknown disturbance is assumed to be deterministic and is represented as an output of a linear autonomous model with unknown constant parameters. To solve the problem of adaptive disturbance compensation, an adaptive observer of an unknown deterministic disturbance is applied. A nonlinear robust control of a multi-DOF elastic electromechanical plant, synthesized on the basis of the integrator backstepping method and combined with adaptive robust disturbance compensation, is also being developed. The simulation of the designed nonlinear control of a multi-DOF plant with adaptive robust compensation of disturbances is performed using the MATLAB / Simulink software.

The paper considers the problems of control synthesis of a rigid multi-degree-of-freedom (multi-DOF) nonlinear mechanical plant under conditions of parametric and functional uncertainty. Developed and investigated adaptive robust control systems are synthesized on the basis of three approaches: based on the method of computed torque, the method of majorizing functions, and also the method of integrator backstepping with modified adaptation algorithms with parametric projection. The results of a comparative study by modeling the designed adaptive systems for the nonlinear mathematical model of a four-degree manipulation arm described by the Lagrange equations are shown.

The paper considers development of an intelligent system for real-time management of groups of small satellites, as well as an approach to its implementation built on the basis of multi-agent technology. It discusses the main disadvantages of traditional planning systems currently used to manage various space missions, which complicate their application with new-generation multi-satellite orbital groups. A brief problem statement for planning the operation of a space observation system is given. The architecture of the developed system is proposed, describing the functions of the main modules and screens, which make it possible to process applications for imaging of observation objects entering the system in real time. Results of experimental studies are also presented. They demonstrate applicability of the proposed approach to solving problems of managing groups of small satellites to increase efficiency of using resources of the space system. Finally, the prospects for development and application of the system are discussed.

A way of constructing a logic ontology for complex structured objects is considered in the paper. The descriptions of such objects are given in the predicate calculus language. An algorithm for extracting subsets of objects with the same properties and establishing their relative position is described. To construct such an ontology, an algorithm for extracting the largest subformula isomorphic to subformulas of all descriptions of objects from the class is used.

The examination of chaotic phenomena origination and their dynamics is a promising area of studies into technical malfunction forecasting. This paper provides analysis of the Lorenz system on a basis of the expansion with the 'Caterpillar'-SSA method. The trend, periodic and non-cyclic components are determined. The necessary number of principal components for time series reconstruction on a basis of Lorenz system is obtained.

This paper provides the trajectory planning and motion control schema for collaborative robot manipulator in a dynamic environment using the computer vision system. The key result is a real-time motion control algorithm using potential fields approach. The simulation results illustrate the performance of the proposed algorithm. As a test platform, the KUKA iiwa manipulator is used.

The purpose of this work is to consider the methodology of designing intellectual models of cyber-physical social systems that take into account the cybernetic properties of the hardware, software and the human factor specifics of the system developers and users. To ensure customizable solutions, model design process is based on systems analysis. This methodology can be implemented by optimizing the objective tree, which determines the development vector of the model being designed based on their importance; the information tree, which determines the structure of the knowledge base optimized by the cost of resources used; the task tree, which allows optimizing the development of the cyber-physical social systems based on the results of analyzing the criteria of the two trees. Intellectual properties of the model being designed are determined by automating the management of its development vector in accordance with the dynamics of the utility of its specific goals. Due to practical limitations on the amount of resources available, the task is to optimize the plan for creating a cyber-physical social system that maximizes the solution value with limited cost. The choice between more precise algorithms that are not applicable given fuzzy limitations caused by human factor allows solving the problems faster than when using deterministic or heuristic methods.

The issues of organizing collective decision-making support for innovative project teams are considered. Research relevance is justified by the need to reduce the degree of uncertainty in decision-making during the projects set implementation, using accumulated corporate experience and knowledge. A synthesis of the subject area and information systems functionality ontological models is proposed. This will improve the efficiency of information system reengineering and the quality of made decisions, among them during the process of collective decision-making support system development.

The paper is concerned with paired tenzotremorogramms that are a result of finger force registration (for both human arms) via sensitive platform with tenzosensor. The main hypothesis is: the dynamics of force hold process differs for different patient category (for healthy patients and for patients who have some pathology). In the paper some approaches to paired tenzotremorogramms structure similarity analysis are investigated. These approaches are based on application of time series distance functions.

The problem of metric identification of probability density distribution for a random value is considered. It is shown that the accuracy is determined both by the sample size and by the number of intervals into which the domain of its determining is divided.

The paper deals with the issues of data collection in the dynamic Internet of Things network. An approach proposed is based on the models and methods of social networks analysis along with clustering methods commonly used in this case. The developed hybrid model tracks contacts between nodes of the network thus creating "social relationships" for each node and estimates the number and the frequency of "social contacts" characterizing the degree of "friendships". After fixing the contacts between nodes the model takes into consideration only the nodes connected by "friendship" relation. Data collection based on "social structure" provides higher security level and shows the enhancement as regards energy saving and delays of data transfer.

An analytical model is proposed for estimating the probability of timely redundant servicing of a heterogeneous flow, which implies the creation of replicas of requests with a multiplicity depending on the total allowable waiting time in the sequence of nodes involved in servicing the request. It is shown that the reservation of queries critical to the admissible waiting time allows increasing the probability of timely execution of queries and the intensity of the profit received from servicing a non-uniform flow.

The amount of stored and transmitted information is constantly increasing. The search and retrieval of the required data become more difficult as it grows. In this paper, we consider the developed storage system. The system stores information about pharmaceutical research. This system effectively stores data depending on its features. The main attention is paid to the issue of data caching at the application server level. To speed up data retrieval, we suggest using a two-section cache. The first section contains the results of queries to the database, and the second contains data from the file system. Analytical and simulation caching models have been developed for research. In the model, you can specify the distribution of requests, the size of the cache sections, the number of requests, and other parameters. The model uses the LRU cache replacement policy. The graphs of model research depending on the cache size and other parameters are provided. The analytical model considers the ideal case, so it shows better results compared to the simulation model.

. It is noted that the manufacturing control optimization is an effective way both in the production processes automation and in the application of the mode of an operational transition to the new products release. Joint consideration of the options from the standpoint of system analysis made it possible to form a methodology of structural-algorithmic modeling, on the basis of which a homogeneous mathematical method is proposed. Interesting results for technical and economic procedures with an aggregate consideration of operational procedures are presented. Results for operational procedures (with detailed representation) together with the theory of description are obtained simultaneously. It is noted that a more complete use of optimization requires a joint description of the specified subject areas. The proposed solution is built on the basis of cyber-physical systems, methodology of structural-algorithmic modeling and a homogeneous method of description. The illustration on a two-block model of the system is given. The features of operational procedures from the standpoint of an integral description are formulated, and a joint description of subject areas is presented.

Significant weight of subjective models used in the systems analysis is discussed. It reduces degree of correctness (coherence and completeness) of planning and managerial decisions as results of the systems analysis in the production sphere and causes costs of production in the decisions realization. That is why a methodological basis for decision-makers objectifying the systems analysis results is so urgent. It is based on the original system-goal approach, logical and linguistic models of goals, semantic model of structure of goals and semiotic model of reasoning about goals. Three main tasks for objectifying the results of systems analysis are set. Schemes, the main procedures, instructions and examples for the task solution are given.

The paper is devoted to the study of processes in the bioreactor with the structural scheme of the University of Cape Town. The ASM1 mathematical model was used to study purification processes. Operating values of recycle flow rates are determined depending on input flow rate required for effective purification of ammonium nitrogen, and a method is proposed to expand the range of control of these values in combination with regulation of oxygen supply to the aerobic zone of the reactor. The specific role of the aerobic zone in the UCT reactor having a key role and affecting both nitrification and denitrification due to nitrate recycling is shown.

The relevance of the network traffic prediction is due to the requirement to store a large amount of data for a long time. In this regard, it is necessary to predict traffic volumes in order to take the necessary measures to protect and preserve data. The paper substantiates the use of autoregressive class models to construct predictive estimates. The results allowing substantiating the prospective requirements for the memory volumes of the node equipment of the infocommunication network are given.

The purpose of the paper is to demonstrate the possibility of improving the quality of logistics in a rapidly changing situation based on mathematical modeling. The evolution of information technology worldwide goes from separate accounting of individual operations to their integration based on cloud and integrated technologies. Adaptive optimization is required, taking into account the dynamics of the external environment. The paper proposes a flexible mathematical model of management transformation in logistics. For this, a single digital logistics platform is formed as a connecting link for all participants in the value-added chain, including product manufacturers, suppliers of resources and services, product consumers, and logistics companies. At the same time, intermediaries who do not create added value could be excluded from the supply chain.

This study aims at the issues of evaluating the effectiveness of organizations and determining significant factors for its improvement. Effectiveness valuation is measured by the distance of the organization to the efficient production frontier (EPF). We consider two main approaches to the construction of EPF: Data Envelopment Analysis (DEA) and the Stochastic Frontier Analysis (SFA). We analyzed and compared these approaches, applying the developed procedures to evaluate the effectiveness of two groups of organizations from different production sectors (mechanical engineering and oil refining). We conducted 1) building empirical estimates of technological efficiency, 2) identifying factors that contribute to the technological inefficiency of the organization, 3) comparing efficiency factors and inefficiency factors for two groups of enterprises.

The problem of transport routing in a regional transport network is considered. The solution method is the Minimum Cost Flow Problem, and the Column Generation Method. For the conditions of degenerate cases, the paper proposes a description of a transport system in the form of directed graphs. As a result, various forms of routes are obtained, as is customary in the transport logistics of road transport.

Aspects of compatibility of topologies of parallel computing systems and tasks are investigated. The introduction of appropriate indexes based on the original topological model of parallel computations and on the nontraditional description of a graph by its projections is proposed and elucidated. On the example of hypercubic computing system (CS) and tasks with ring and star information topologies, we demonstrate determining the indexes and their use in a comparative analysis of the applicability of interconnect with a given topology for solving tasks with the same and different types of information topologies.

The thesis of the futurologist A. Clark that «sufficiently advanced technologies are indistinguishable from magic» is relevant nowadays as never before. Modern science, unlike magic, although it relies on logically consistent physical theories, formalisms of mathematics and the possibilities of computing, has not made the world less mysterious, especially with regard to the natural phenomena of consciousness, thinking and intelligence. From the point of view of computer science, a distinctive feature of these phenomena is their network-centric neuromorphic organization and the availability of complex memory resources - a non-casual computing mechanism for «transporting» processed data, the results of calculations and information interactions in time. Formalization of these features makes it possible to determine a model of cognitive processes on a multidimensional set of states, on which a strict relation between the time order and numerical equality is not introduced, therefore, contextually significant information, about past as well as current or forecast states, can be used for calculations at the same time states. For control systems operating under uncertainty, a distributed heterogeneous reconfigurable structure is proposed that implements the technology of «computation in memory» both for algorithms reflecting causal relationships and learning processes that require operational reconfiguration of the computation field.

The problem of ensuring the stable functioning of time series forecasting systems based on streaming recurrent neural networks with controlled elements is considered. The mechanisms necessary and sufficient for its maintenance are derived, which involve maintaining the balance of the learning history and modifying the synapse learning rules in order to establish a balance between positive and negative potential. The results of experiments to assess the accuracy of forecasting are presented.

Methods of accuracy improving of pre-trained networks are discussed. Images of ships are input data for the networks. Networks are built and trained using Keras and TensorFlow machine learning libraries. Fine tuning of previously trained convoluted artificial neural networks for pattern recognition tasks is described. Fine tuning of VGG16 and VGG19 networks is done by using Keras Applications. The accuracy of VGG16 network with fine tuning of the last convolution unit increased from 94.38% to 95.21%. An increase is only 0.83%. The accuracy of VGG19 network with fine tuning of the last convolution unit increased from 92.97% to 96.39%, which is a 3.42% increase.

A stratified model for self-similar modular neural networks that includes morphological, structural, topological and parametric levels is proposed. It is shown that the morphogenesis of the morphological model is determined in the population of terminal projections of the neural network modules. Structurally regular neural networks are considered. Simplification of modular neural network models without loss of functionality is achieved using translational connections. It is shown that fast transformation algorithms (including FFT) can be described by a topological model of a structurally regular self-similar network. A linguistic model for describing the topologies of regular self-similar networks is presented. An algorithm for constructing topological models of fast algorithms is proposed. The sufficiency of the topological model for describing the complete set of fast algorithms is shown. Parametric model of modular neural network is presented. Examples are given.

In this work, we develop a systemic approach to the study of a new model of COVID-19 pandemic. The main goal is to minimize the pandemic damage to economy and society by defining the model optimal management parameters. Our approach consists of two main parts: 1) the adaptive-compartmental model of the epidemic (ACM-SEIR) – a generalization of the classical SEIR model and 2) the module to tune ACM-SEIR parameters using artificial intelligence methods (collection, storage and processing of big data from heterogeneous sources) that allow the most accurate adjustment of ACM-SEIR parameters turning it into an intelligent system for decision support called herein iACM-SEIR. We show that among iACM-SEIR parameters, the most important are individual economic, demographic and psychologic characteristics of society and the governmental actions.

Smart contracts created to issue tokens and control other digital financial assets on the majority of blockchain platforms require knowledge of programming language for specific platforms. This fact imposes a restriction on attracting financial specialists from conventional markets. This paper presents the concept of a Restricted Move language allowing the users to form a smart contract description in financial English, as well as its further development into a high-level digital asset management language, which can be extended by users and can open new opportunities to automate their market operations.

The report is devoted to the study of the problem of parametric identification of controlled dynamic systems in their normal operation mode using a model with adjustable parameters. The task of determining parameters of test signals providing the specified accuracy of identification is considered. The peculiarity of the work consists in imparting adaptive properties to the identification procedure, which makes it possible to minimize the negative effect of "swinging" of the controlled object on its output when special test signals are applied.

Forest is the most important natural resource of the Baikal natural territory (BNT), which ensures sustainable development of the ecological system of the Lake Baikal, performing water protection, soil protection and water regulation functions. The article describes the issues of transition to a new technological order - the creation of a digital monitoring platform to support the rational use of forests. We formulate characteristics of the current state of forests and analyze the problems of forest resources monitoring in the BNT. We consider the types of digital platforms of different levels and propose basic requirements for forest digital monitoring.

In this work we analyze and compare machine learning methods for recognizing human activity in the context of smart home by using data obtained from an optical heartbeat sensor and an accelerometer embedded in a smart watch, and find a number of activity classes that can be predicted. The conclusion is made that for such type of problems the random forest method with about 10 classes shows the best results.

. In this paper, complementary objects design and usage is considered. The presented examples illustrate the programming paradigms necessary for correct usage of complementary objects that are essential in distributed computation systems.

The report is devoted to the applications of the envelope method to the synthesis of control in the differential game "pursuit-evasion". It is proved that optimal trajectories can be represented as envelopes of a parametric family of singular curves, called instantaneous solutions within this approach, and that control can be found on this family. It is shown that, in the general case, the synthesized game control law will contain the maximum permissible control values, current values of phase coordinates, parameters of tangents and their derivatives.

The traditional traceability system for agricultural product transactions is susceptible to information alteration and damage, which may lead to issues regarding product quality and food safety. And blockchain is a technology that boasts tamper-proofing, complete traceability and time-stamped storage. Considering the above, this study proposes a new blockchain-based approach to the quality management of agricultural products and introduces Solidity-based prototype smart contracts for agricultural product transactions. Test results show that the new traceability system can offer good performance in terms of data upload and block response time. The method proposed in this paper can be used as a solution for quality management of agricultural product that boasts whole process transparency, full-link reliability and joint supervision by all nodes in the system.

The paper presents a method to estimate the frequencies of a multi-harmonic signal in finite time. We use parameterization based on applying delay operators to a measurable signal. The result is a linear regression model with an unknown vector which depends on the signal parameters. We use Dynamic Regressor Extension and Mixing method to replace the n-th order regression model with scalar regressions. After that, we estimate the parameters separately using the standard gradient descent method. In the last step, we find algebraically the finite-time parameter estimates. The set of numerical simulations demonstrates the efficiency of the proposed approach.

In this paper, a comparative analysis of methods for automatic classification of text information using various machine learning algorithms based on artificial neural networks is carried out. The paper presents results of the text classification considering various ways of a vector space formation, such as bag-of-words, n-gram, word embedding, and various architectures of neural networks. The research is based on the text corpus in Russian.

The process of frame development for Multipurpose Control and Display Unit (MCDU), Multi-Function Display (MFD), and Primary Flight Display (PFD) involves a significant documentation effort and must produce consistent interface at the early design stages. We developed a formal description of the interface, including the interaction vocabulary, as well as custom software to simplify these tasks and improve the overall design quality. The software builds a model of the human-machine interface and uses it to automatically generate frame specification/documentation. This methodology and the software are currently used in several commercial projects.

This paper suggests improving the previously existing method of identifying user profiles in different online social networks by adding face recognition results to the model. It is assumed that the method will become more stable for identifying people with the same name, city and age. It will help to find more user profiles in different online social networks, which will improve the estimation of their personal characteristics. Evaluating user personality traits is one of the key tasks in protecting employees of enterprises and companies from social engineering attacks.

The problem of ensuring the information security of the smart home is discussed, the main vulnerabilities and the consequences they can cause are listed. Mechanisms for protecting information in the smart home system based on the characteristics of consumed resources: energy, time, computing power, are proposed. A description of the smart home system assembled model is presented, and consumed resources characteristics allowing reasonable choice of the method for protecting information transmitted by devices over open channels are measured by using this model.

A technique for classifying the social network users and groups by psychological scales of the Ammon's test has been developed. To analyze user profiles, we have used several types of artificial neural networks (support vector machine, linear regression, multilayer neural network and convolutional neural network), and for group analysis, we applied text classifiers (bag of words, weighted bag of words, continuous bag of words, skip-gram and fastText classifier). The scope of the technique is identifying deviations in the psychological state of users of social networks and monitoring these changes considering users' groups to detect destructive influences. An experiment was carried out, as a result of which it was found that a multilayer neural network with an activation function of the ReLU type has the best accuracy.

In the process of monitoring industrial facilities information security, the sets of heterogeneous data are generated by different monitoring tools which are to be processed according to various regulatory documents not always compatible with each other. Definition of a set of security characteristics based on perimeter monitoring data and determination of security level carried out mostly by calculating integral indicator do not take into account some aspects of security. The method is proposed for the harmonization of industrial facilities interaction and for the formation of a set of security indicators on the basis of formalized subject domain description and unified data description scheme.

The paper deals with a topical subject of information system access security through development of a polling system aimed at forecasting the resistance of the user's passwords against brute force. A method of development of such a system, which is based on computer-aided learning, is suggested. This method is based on the following hypothesis: "Social and psychological features of the password system users have a bearing on the complexity of created passwords." The method is elaborated in the form of the following 6 steps: development of a group of questions which reflect the preferences of the password system user; implementation of a polling system which requests the user's typical password; selection of a mechanism of independent estimation of password complexity; application of the polling system to generate implicit interrelations of specific features of users and their passwords; development and training of a neural network to solve a regression problem based on the results of the polling system use; and estimation of the neural network by a test set of users. The experiment results are presented in a form of a diagram. The results of the experiment are discussed and conclusions are drawn about the method development.

The current level of production development requires the training of qualified specialists capable of solving complex scientific and technical problems in the field of technical systems management using rapidly changing information technologies. Consequently, the training of professional personnel should provide for the needs of developers of advanced control systems for complex dynamic objects. This problem is successfully solved at the Department of Automation and Control Processes (ACP department) in ETU "LETI". Students training is carried out in two complementary directions "Management in technical systems" and "Information systems and technologies". Long-term experience proves the advisability of choosing such an approach in the context of rapidly developing technologies. The report contains studies based on the application of data analysis methods and models that dynamically integrate the information necessary for adaptive management of the process of advanced specialists training. The dataset includes: actual employers' data, production needs, on the one hand, data obtained during the educational process, on the other hand. At the same time, the data set includes data that enable students to assess the effectiveness of the educational services provision, the level of professional training in the areas of specialty. Comprehensive analysis of data integrated in a unified environment, based on the methods of intelligent systems, analytics and knowledge engineering, proves the validity of choosing this approach in the learning process.

The Orlan (Russia) and the Pelican Ultra (USA) ekranoplane projects are analyzed and compared in terms of their design features and applications. Particular attention is paid to the challenges the designers faced while creating ekranoplanes and their control systems. The originality of the tasks that have to be solved by modern and advanced control systems of ekranoplanes, is considered.

The article deals with the theory of designing ekranoplane control systems, taking into account the features of their dynamic properties and operating conditions. Flying at low altitude over an uneven surface is accompanied by powerful periodic and pulse loads caused by the roughness of the underlying surface. In addition, with small changes in altitude, the controllability and stability of the ekranoplane changes significantly up to loss of stability. For the synthesis of the ekranoplane control system, a new method for calculating a nonlinear mathematical model of an ekranoplane based on the CFD Comsol Multiphysics and Matlab programs is proposed. The results of simulation are presented.

This paper presents the assessment of cybersecurity of the non-industrial unmanned aerial vehicle (UAV; using DJI Tello as an example) wireless control channel, implemented on the basis of Wi-Fi Wireless Local Area Network technology. As a result of the research, a model of an adversary conducting a cyberattack on the UAV communication channel was developed, and scenarios of an intruder's actions (aimed at taking control of the UAV or disabling it) were considered. Relevant block diagrams of the actions of an adversary were also developed. Based on the results of the analysis, a security strategy is proposed. It provides the maximum possible protection (based on the technical characteristics of the drone) and comprises not only passive protection in the form of setting a tough password, but also active actions –deploying a system to monitor the injections of deauthentication/dissociation frames and/or the creation of duplicate wireless networks.

The article aims to develop a neural model for digital air traffic control. The proposed approach uses the concept of socio-cyber-physical self-organization of distributed organizational and technical system, whose components are connected to 4G and 5G generation wireless networks. The advantage of the approach is complex integration of promising analytical management principles and their representation in hybrid intellect.

During deep space missions, planning and realization of dynamic operations (orbit and attitude correction manoeuvres, landing in target areas, etc.) require fast determination of spacecraft position. Due to the objective difficulties of interplanetary communication, it is interesting to solve this problem using the data from onboard sensors (in particular, the optical ones). What is more, most "transportable" of these devices give out "angular" (or quaternion) representation of reference objects' positions instead of rectangular coordinates or at least distances. Thus, one faces the problem of determination of spacecraft state vector (in a rectangular coordinate system) from the data in "angular" coordinates in the coordinate system of measurement devices. We suggest to determine the spacecraft state vector at the required time as a solution to the problem of minimizing the quadratic cost that determines the difference between the "simulated" measurements (i.e., those that would have been made on the "simulated" trajectory generated by a current state vector approximation) and the real ones (i.e., those that were obtained from on-board sensors). To find a solution, we use a software implementation of the Levenberg-Marquardt method and a prototype of a software package that allows one to configure spacecraft sensors set and the measurement series parameters.

The problem of optimal control of the spacecraft interplanetary flight from the Earth to Mars with a three-pulse approach to the Phobos tracking orbit is considered. The departure of the spacecraft from the Earth is also approximated by pulses. The attraction of the Earth and Mars is taken into account. The problem is solved numerically on the basis of Lagrange principle. The results are compared with the scheme of a full turn free approach to Phobos. The initial approximation is carried out based on a combination of Lambert problems solutions. The software package is developed. Ephemerides are taken into account in the calculations, and the NASA SPICE package is used.

The paper investigates the service satellite control system when approaching a non-cooperative spacecraft in the state of uncontrollable motion. The service satellite control system is considered as a two-level multichannel control system, and multicriteria optimization methods with cross-links are developed. The mathematical model of the control system for a free-flying space manipulation robot at the guidance and stabilization phase is proposed. The efficiency criteria are presented.

Methods for the motion control of a space robot flying by an information geostationary satellite during visual inspection of its technical state are considered. The robot attitude and orbit control system employs electric propulsion unit with 8 catalytic propulsion engines and pulse-width modulation of their thrust, and also a gyro moment cluster based on 4 gyrodines with digital control. The results of computer simulation demonstrating the effectiveness of the developed discrete control algorithms are presented.

The paper discusses the motion cues simulation problem for aerospace applications. An algorithm was suggested for controlling a motion stand based on an industrial robotic manipulator. The paper provides simple mathematical model reflecting the movements of manipulator's endpoint in the vertical plane with the base performing rotational movements and presents a restriction set in the configuration space. We provide an algorithm for manipulator's stopping and returning to the start position of a motion simulation process active phase. A set of restrictions for the motion simulation active phase is also obtained.

Intelligent data analysis systems are designed to minimize the efforts of a decision-maker in the process of data analysis. The article deals with the tasks of intelligent processing and analysis of data from an electronic engine control system unit during testing. An intelligent data analysis system is proposed with the subsequent decision-making on the states of the engine and the electronic control system during test flights. With the help of artificial intelligence methods, causal relationships and regularities of the parameters of a turbojet engine have been revealed.

A numerical method for calculating the generalized additional masses of sea underwater objects with fins, performing both single and joint navigation of several objects, as well as with elastic vibrations of the hull, is presented. The flow is modeled using a continuous layer of sources and vortex frames that simulate the body and its fins. The method provides a strict account of the mutual influence of bodies and their elements and does not impose restrictions on their shape. Approbation of the method showed its reliability and the possibility of practical use in design practice, as well as in simulators and on-board complexes.

The work is devoted to the problem of critical situations recognition in real-time and localization of increased acoustic vibration in critical objects, based on the signals from segmental distributed fiber-optic measuring networks (DFOMN). To solve the problem, an architecture is proposed, and a model of a situational approach to using the Convolutional Neural Network as an effective classification method for processing large data arrays of the DFOMN is developed.

The indicators of the quality of devices for trajectory processing of radar information and methods of their testing are considered. Practical recommendations are proposed that simplify the assessment of quality indicators of trajectory processing devices.

The general principles of construction and features of the operation of the trajectory processing device based on data from one source of radar information are considered. The tasks performed and structure of the trajectory processing device are discussed.

The possibility of using the Basis Pursuit signal recovery method in the object tracking algorithm as a method of filtering and extrapolating the trajectory of the observed object in a one-dimensional measurement space is considered. The simulation results are presented for the trajectory processing algorithm using the proposed filtering method for the data obtained during the object direction-finding in comparison with the block algorithm based on the maximum likelihood method and the Kalman filter.

The problem of synthesizing a direction finding algorithm based on a sample composed of the output signals of a limited number of channels of a scanning sonar system is considered. The problem is formulated as a search for the maximum of the likelihood function by its lattice representation. Options for using the optimal and quadratic interpolation are considered. The errors of displacement and loss of potential accuracy are estimated depending on sample dimension, angular step of scanning, and on the amount of the target movement within the measurement interval. The advantage of the proposed approach is shown in terms of its potential accuracy compared to the traditional approach to constructing direction finding paths based on Guanella's technique.

The ways of overcoming the problems caused by the specific asymmetry of the side lobes of the channels formed at the output of cylindrical antenna arrays with a horizontal generatrix, which are relevant for underwater robots, are considered. An algorithmic solution is proposed based on the use of adaptive signal power limitation for solving detection-direction finding problems in combination with non-adaptive reception or the diagonal loading Capon algorithm for assessing the spectral and correlation characteristics of strong signals.

The paper considers an approach to the construction of a clutter model used in software of the technological part of the test bench for simulation and development testing of hydroacoustic systems. Simulation of clutter on antenna arrays is a rather complex independent task important from the point of view of ensuring the correct SNR for the simulated noise sources at the antenna input. The existing methods of clutter simulation are presented and a new approach is proposed that eliminates the existing drawbacks while maintaining high performance.

The problem of coordinated operation of a hybrid system for generation and distribution of electric power to provide motion for an underwater vehicle is addressed. The architecture of a multi-agent vehicle system is chosen, the optimum distribution of consumers to electric lines is defined, effective switching of consumers between electric lines, with due regard to the functional and non-functional requirements for the system, is developed.

Practical development of easy-to-implement technical solutions for receiving and transmitting a surface electromagnetic wave (SEW) – that passes through the skin layer and propagates within the boundaries of the sea surface in the lower troposphere and the water's edge – from a submerged position will create preconditions for the use of radio control channel in marine systems. The report reveals the topic of SEW using for radio control of unmanned underwater devices and describes the results of design work on the ultra-low radio frequency underwater small-sized antennas.

The report presents a comparative analysis of various control allocation methods by taking into account an over-actuated propulsion system and actuator constraints for autonomous underwater vehicles. Comparing methods by the size of a feasible control region, compute time, consumed power, and control error is presented. It is shown that the choice of an allocation method significantly affects the quality of control of an underwater vehicle and consumed power during maneuvers.

A range of issues and tasks related to the construction of a measurement system based on integrating fiber-optic sensors for monitoring and controlling low-frequency acoustic vibrations of ship hull structures is considered.

The paper presents a numerical method for calculating the hydrostatic and hydrodynamic reactions of marine underwater objects moving in an arbitrary manner under conditions of internal waves and sea currents. The method is based on the theory of potential, eddy and wave currents. The effect of viscosity on hydrodynamic reactions is taken into account indirectly by introducing a special eddying system that simulates the effect of longitudinal vortices on the hull on the pressure distribution. The paper presents the results of a computational experiment are presented, which made it possible to identify both new and confirm previously obtained regularities of the influence of the object's parameters and hydrophysical conditions on the desired reactions. The developed mathematical and software tool can be recommended for use in the design of underwater objects and the creation of simulators for their dynamics.

The report proposes an approach to assessing the minimum reliability level of control panel operators of the ship nuclear power plants based on a combined probability-deterministic analytical-simulation method, functional-structural theory and probabilistic-algorithmic models of the ergatic systems functioning.

The possibility of predicting the cost of construction of a new type of transport vehicle based on analogies with already developed related areas of transport engineering is discussed. Special attention is paid to assessing the complexity of creating ekranoplanes and their control systems. Features of the Arctic application of heavy ekranoplanes are studied in comparison with heavy icebreakers. The originality of tasks that should be carried out by modern and promising control systems is evaluated.

Modeling of full-scale propellers characteristics at the design stage is one of the actual practical tasks. A method has been developed for converting characteristics from the model to nature to solve this task. The method is based on modeling the propeller operation using Eduard Papmel modified method. The boundary layer characteristics are calculated for the blades cylindrical sections and the axisymmetric body modeling the hub. Compliance between the evaluation characteristics and a model tests result is ensured by selecting a number of parameters that take into account the features of the flow around the blades and the incoming flow.

The necessity of improving methods for determining the technical condition of ship power plants is justified, using the example of an auxiliary engine. A brief analysis of the most common systems of self-diagnostics is given. Features of development and use of artificial intelligence for diagnostics of internal combustion engines are described. The algorithm for conducting experimental studies to obtain a sample of diagnostic features is incorrect. The characteristics of synthesized neural networks are given and the accuracy of determining the technical state is estimated.

A concept is suggested for modeling the dynamics of the fit of a displacement river vessel trimmi for making decisions about the readiness to use technical means of evacuation and rescue of passengers and crew in the event of a rapid flooding threat. The concept is based on the imitation of loading a large cargo, prone to displacement. The concept is implemented through continuous monitoring of the vessel fitting dynamics in a beyond design basis emergency, real-time forecasting of its development, and the captain taking proactive measures to evacuate passengers and crew.

The paper considers the problem of dynamic positioning nonlinear model of a vessel under the multi-harmonic disturbances. The multipurpose control law makes it possible to tune the nonlinear observer, the feedback control, the dynamic corrector, and the estimator of external disturbances frequencies independently from each other. We achieve the economical ship control mode by tuning the dynamic filter to the primary external perturbation frequency. We apply a finite time approach to construct estimates of multi-harmonic external disturbance frequencies.