Category Archives: Technical systems

TECHNIQUE FOR A QUICKLY DEPLOYED COMMUNICATION NETWORK CREATION BASED ON RADIO RELAY LINES


UDC 621.396.4
DOI:10.26102/2310-6018/2020.30.3.032

V.O. Klyuchnikov

Radio relay communication plays an important role in the creation of digital infrastructure. Modern digital radio relay lines allow organizing high-speed communication channels supporting the transmission of all types of traffic, including voice messages, packet data transmission, and video signals in real time. The article discusses methods for calculating digital radio relay communication lines, and also provides general recommendations for organizing digital radio relay communication lines. The design of radio relay lines is carried out in such a way that they have high bandwidth and high reliability, taking into account the physical (energy) and channel-network capabilities of the system. The calculation method requires determining the energy parameters, deriving the form of the quality equation, calculating the performance criteria and methods for calculating the attenuation of radio waves at radio relay intervals. The possibility of using MPLS-technology in the radio relay communication was positively assessed. The radio relay network routing system is based on multiplexing streams with a protection mechanism at the channel and network levels. Continuous adaptive modulation mode allows flexible QAM rate change, allowing continuous traffic. The redundancy according to the «1+1» and «2+0» schemes is considered. This scheme doubles the system throughput and prioritizes traffic. The technique allows you to build a digital radio relay network with flexible changes in both physical and network parameters.

Keywords: radio relay network, radio relay lines, calculation methods, quality indicators, radio wave propagation, packet delivery route.

Full text:
Klyuchnikov_3_20_1.pdf

SIMULATION MODEL OF THE RADAR ENVIRONMENT OF AN INTELLIGENT CONTROL SYSTEM FOR DISTRIBUTED RADAR FACILITIES


UDC 004.942
DOI:10.26102/2310-6018/2020.30.3.038

M.B. Sergeev, A.A. Sentsov, E.K. Grigoriev, S.A. Nenashev

The beginning of the mass use of small unmanned aerial vehicles for various purposes gave rise to the problem of their safe and controlled movement in space. Accordingly, the challenge arises of constructing new intelligent control systems. In the approximate list of actions that these systems should carry out, you can specify the control of the operating modes of each data source for scanning the airspace, obtaining information about moving objects, calculating the coordinates and direction of movement (component velocity vectors) from the processed data, as well as predicting the position of the air object to decide on the issuance of information to paired systems. The article describes the modeling process of a distributed system consisting of two mobile radar stations, applicable for developing methods for detecting and estimating the coordinates of airborne objects. For the simulation model being developed, analytical relationships are obtained for calculating the coordinates of the observed physical objects using both range-finding information and goniometric. A block diagram of the modeling steps for determining the trajectory coordinates of the object is proposed. The expediency of using a distributed system in order to improve the accuracy of measuring trajectory coordinates is shown. Various options for the location of spaced observation points are proposed, as well as their advantages and disadvantages. The model is built taking into account experimental data obtained from a simulator of an air moving object. The developed simulation model is designed to conduct experiments, as well as to determine the performance characteristics at the stage of development of functional interactions of the intelligent control system for distributed mobile devices.

Keywords: determination of coordinates, air object, two-position system, radar system, complex information processing, simulation model.

Full text:
SergeevSoavtors_3_20_1.pdf

TWO-STAGE PROCEDURE FOR THE SYNTHESIS OF CONTROL OF NONLINEAR NON-STATIONARY OBJECTS USING A MULTILAYER PERCEPTRON


UDC 62-54
DOI:10.26102/2310-6018/2020.30.3.028

S.V. Frolov, S.V. Sindeev, A.A. Korobov, K.S. Savinova, A.Y. Potlov

The review of neurocontrol methods and analysis of their advantages and disadvantages is presented. The problem of searching of quasioptimal tuning parameters of neurocontrol for nonlinear non-stationary objects in the presence of random disturbances is formulated. A procedure for the synthesis of control for nonlinear non-stationary objects using a multi-layer perceptron, which consists of two stages, is presented. In the first stage the problem of finding a robust neurocontrol vector tuning parameters for adaptation algorithm based on the proposed set of variants of the model is solved. Founded tuning parameters for adaptation algorithm are used in the second stage – model-free neurocontrol, which searching for quasi-optimal tuning parameters for the algorithm of model-free neurocontrol. Stability of tuning parameters search procedure for the algorithm of model-free neurocontrol achieved by using the regularization method. Effectiveness and stability of the proposed procedure for the synthesis of control for nonlinear non-stationary objects are shown using the model example. In the numerical experiment, an object was chosen that was described by a nonlinear differential equation with coefficients that depend on time. At the first stage, 20 variants of the object model were randomly generated, the architecture of the neural network, the tuning coefficients of the adaptation algorithm were found. The neural network includes 2 neurons in the inner layer and uses a sigmoidal activation function. At the second stage, numerical studies of the adaptive control process were carried out. As a result of the adaptation algorithm, the degree of attenuation of transient processes exceeds 50% and the control process is stable with a significant deviation of the object’s parameters from the nominal values. The presented method is effective for the control of multiply connected non-stationary nonlinear objects in robotics, transport systems, and chemical industries.

Keywords: neurocontrol, neurocontroller, multi-layer perceptron, control system, adaptive control.

Full text:
FrolovSoavtors_3_20_1.pdf

3D INDOOR POSITIONING SYSTEM BASED ON BLUETOOTH BEACONS


UDC 004.8
DOI:10.26102/2310-6018/2020.30.3.023

V.M. Grinyak

This paper devoted to research of indoors navigation problems under poor or insufficient quality of satellite navigational data environment. The problem of object positioning in 3D space by Bluetooth devices located indoors forming a multi-position tracking system is considered in this research. Emphasized that in order to succeed for such system it is required to pre-estimate distinctive accuracy. The proposed model interpretation of the positioning problem as the system of linear equations. The classic model interpretation for method of least squares is used for resolution. General problem of linearization around reference resolution is the locality of its features. There are three concepts of problems solvability, such as fundamental solvability (observability), solvability in conditions of instrumental measurement errors and solvability under conditions of finite accuracy of computation on a computer. The first aspect of solvability is interpreted by the completeness of the rank of the corresponding system of linear algebraic equations, the second and third ones represents by the conditionality of the problem and the convergence of the iterative estimation procedure. The conducted experiments show that for the positioning problem the attributes of the linearized model are accurate enough to represent the original nonlinear problem. Such interpretation allows to build theoretical accuracy estimation priors for object coordinates evaluations and to identify the areas with insufficient positioning accuracy. In this paper there are results of expected accuracy evaluation for various system patterns with full-scale experiments proving the theoretical calculations. Experiments for problems with using SKYLAB Beacon VG01 Bluetooth transmitters and smartphone HUAWEI WAS-LX1 are presented and confirmed that math model with linear approximation defined by authors is usable for solving indoors navigation problems using Bluetooth signal. So, for good enough quantity and appropriate location of the tracks the achievable positioning accuracy could be as good as 0.2-0.3 meters for all three coordinates. Such accuracy allows to navigate small hovering objects such as drones. In general, it looks promising to use Bluetooth trackers for solving positioning problems for indoors environments.

Keywords: information system, indoor positioning, beacons, Bluetooth, location, least squares method.

Full text:
Grinyak_3_20_1.pdf

CURRENT-VOLTAGE CHARACTERISTIC OF NON-STATIONARY 1:1 SALT ION TRANSPORT IN THE SECTION OF DESALINATION
CHANNEL


UDC 519.87+004.421
DOI:10.26102/2310-6018/2020.30.3.020

I.V. Shkorkina, N.O. Chubyr, V.A. Gudza, M.Kh. Urtenov

The main goal of this work is to derive and analyze different formulas for calculating the current-voltage characteristic (CVC) of non-stationary transport of 1: 1 electrolyte in the cross-section of the desalting channel, including anion-exchange (AEM) and cation-exchange (CEM) membranes, and to establish fundamental regularities of changes in the CVC with time. Modeling is carried out based on the Nernst-Planck-Poisson equations. The transport of ions of strong electrolytes NaCl and KCl through a thin reaction layer of ion-exchange membranes in the section of the desalination channel is considered. For this purpose, a schematic electrical diagram of the current flow in the circuit, including the cross-section of the desalination channel, has been constructed. From the analysis of this circuit, it follows that the total current consists of a conduction current and a displacement current. The conduction current is determined by the flow of salt ions. The displacement current goes to the formation and development of the space charge region. Due to the change in the increase in the potential jump (potentiodynamic mode), the total current in the circuit when calculating the CVC changes over time, and its change can be considered slow. In this case, the displacement current practically does not pass through the cross-section of the desalination channel while the charge distribution density is slowly changing. In the case of a rapid change in the charge distribution density (breakdown phenomenon, as well as before and after breakdown), the displacement current takes on rather large values. The displacement current-voltage characteristic must be taken into account separately. Since the value of the CVC calculated in the study of the transport current is much higher than the value of the CVC at the displacement current, the effect of the electric “breakdown” on the transport current is hardly noticeable. Therefore, the “breakdown” effect must be investigated by the CVC of the displacement current. The proposed formula for calculating the CVC of the conduction current is stable with respect to rounding errors. The effect of nonstationarity is investigated at high growth rates of the potential jump.

Keywords: current-voltage characteristic, membrane systems, cross-section of the desalting channel, mathematical model, migration current, diffusion current.

Full text:
ShkorkinaSoavtors_3_20_1.pdf

LOW RANK APPROXIMATIONS FOR NEURAL NETWORKS


UDC 004.89
DOI:10.26102/2310-6018/2020.30.3.018

N.V. Shaposhnikova

Today, artificial neural networks (hereinafter ANN) and deep learning have become almost indispensable in applications related to the tasks of machine vision, machine translation, speech to text conversion, text rubrication, video processing, etc. However, despite the presence of a number of classical theorems substantiating the approximating capabilities of neural network structures, the current successes in the field of ANNs in most cases are associated with the heuristic construction of the network architecture applicable only for the specific problem under consideration. On the other hand, deep ANNs have millions of parameters and require powerful computing devices for their functioning, which limits the possibilities of their application, for example, on mobile devices. Significant progress in solving these problems can be obtained using modern powerful algorithms of low-rank approximations for the parameters of the ANN layers, which will both simplify the process of developing a neural network architecture and will lead to significant compression and acceleration of the training of deep ANNs. Considering, for example, the core of the convolutional ANN as a four-dimensional array (tensor), we can construct a low-rank approximation for it with the effective implementation of its convolution with the vector (direct signal propagation in the network when generating the prediction) and differentiation with respect to the parameters (back signal propagation in the network when training). In this paper, we will consider the modern paradigm of machine learning and low-rank tensor approximations, and we will demonstrate the prospects for the tensorization of deep ANNs using a specific model numerical example corresponding to the task of automatic recognition of handwritten digits.

Keywords: machine learning, neural network, deep convolutional network, low rank approximation.

Full text:
Shaposhnikova_3_20_1.pdf

MOTOR CONTROLLER OF AN AUTONOMOUS MOBILE ROBOT WITH NEURO-FUZZY CONTROL


UDC 621.865.8
DOI:10.26102/2310-6018/2020.30.3.012

Han Myo Htun, Alexey N. Yakunin, Htet Soe Paing

Currently, an autonomous mobile robot (MR) is an artificial intelligent vehicle that can bypass obstacles and move to a given point along a predetermined route. In the external environment, one of the main problems is the implementation of a mobile robot on wheels, which moves from one point to another with a detour of obstacles. Such robots are equipped with sensors or a camera, so the MR should be able to detect emerging obstacles. This article proposes the principles of constructing an intelligent controller based on ASNL (adaptive system of neuro-fuzzy logic) of an autonomous mobile robot, allowing to achieve the goal along a predetermined route. The mathematical model of MR and the developed controller in the environment of Matlab-Simulink is implemented. Comparison of the proposed controller with a known fuzzy controller is performed according to the following criteria: path length (DT) and curvature parameter (PC). In this article, the simulation results show that the proposed ASNL controller reduces DT and PC compared to a controller with fuzzy logic (NL), therefore it has better performance indicators. A mobile robot using the proposed ASNL controller is capable of moving toward a given target and avoiding random obstacles without collisions in its path.

Keywords: controller, adaptive system of neuro-fuzzy logic (ASNL), undefined route, mobile robot (MR), fuzzy logic controller (NL), distance traveled, curvature parameter.

Full text:
HanSoavtors_3_20_1.pdf

METHOD FOR JOINT CALIBRATION OF INERTIAL SENSORS OF AN UNMANNED AERIAL VEHICLE USING NEURAL NETWORKS


UDC 629.7.05
DOI:10.26102/2310-6018/2020.30.3.007

V.A. Smirnov, M.N. Pravidlo, A.B. Snedkov

The article deals with methods and technical means of control and testing samples of information-measuring and control systems intended for installation on unmanned aerial vehicles. The principles of creation of control systems built on inertial sensors of various types are considered. Shortcomings are revealed in the use of traditional methods of development based on direct readings from sensors with the help of software. The design of modern inertial sensors with indication of their components and parts is described. The technology of creating a test bench for evaluating the accuracy of sensor calibration is defined. Formulas on which the work of the test bench is based on its geometry are given. Specifies formulas used in the traditional method of calculating values from inertial sensors. Describes the process of learning the neural network to compensate for the disadvantages inherent in a traditional calibration method. The experiment of comparing actual values with the values obtained during the calculations is demonstrated. The traditional method of joint inertial sensor calibration is compared with the method based on neural networks. The advantages of using this method of inertial sensor calibration under the requirements of modern control systems are determined.

Keywords:sensors, gyroscope, accelerometer, calibration, control system.

Full text:
SmirnovSoavtors_3_20_1.pdf

APPROXIMATION OF EVOLUTIONARY PROCESSES WITH DISTRIBUTED PARAMETERS ON A NETWORK


UDC 517.927
DOI:10.26102/2310-6018/2020.30.3.003

O.R. Balaban

The paper considers the approximation of mathematical models of network-like evolutionary transport processes as applied to differential systems with distributed parameters on a network (graph). An approach is indicated that uses the application of the theory of classical computational methods, which consists in reducing the investigated problem to systems of algebraic equations (auxiliary finite-dimensional problems) in which the values of the grid functions at the points of partition of the edges of the graph are unknown. At the same time, there is a fairly wide opportunity for choosing different types of convergent difference schemes that are significantly different from each other: explicit difference schemes, implicit difference schemes, analogs of Krank-Nicholson difference schemes (below, in order not to load the study with technical difficulties, explicit difference schemes are used). It should be noted a characteristic feature of the studied mathematical models inherited by the rheological structure of the graph — the presence of singular points of the graph at which the differential equation is not determined (nodes or vertices of the graph) and is replaced by generalized Kirchhoff conditions. The formalisms of the latter describe the laws of continuum transfer at these points and require a separate approach to approximation issues (in the work, for the sake of simplicity, classical difference relations are used). It should also be noted that the use of an implicit difference scheme or the Crank-Nicholson scheme for approximation requires additional analysis of auxiliary finite-dimensional problems (solvability, uniform boundedness of approximations to the solution of the original problem), but it significantly increases the accuracy of calculating approximations. The use of an explicit difference scheme is freed from studying some of these issues, however (and this, when analyzing some applied problems, can be a significant obstacle to use) gives a rather large error in determining the solution to the original problem. The given particular examples of applied character illustrate the ways of numerical analysis of differential systems with carriers on an arbitrary network (graph). The results obtained are quite simply transferred to the study of wave processes and oscillation phenomena in transport processes by numerical methods.

Keywords:evolutionary transport processes on networks, approximation, difference scheme, features at network nodes, numerical methods.

Full text:
Balaban_3_20_1.pdf

METHOD OF COUNTERING DESTRUCTIVE ELECTROMAGNETIC INFLUENCES, BASED ON ADDITIONAL MODULATION WITH APPLICATION OF WAVELET TRANSFORMATION IN SPECIAL APPLICATION NETWORKS


UDC 654.16
DOI:10.26102/2310-6018/2020.29.2.039

I.V. Gilev, S.V. Kanavin, A.V. Popov, N.S. Khokhlov

The article discusses a method of counteracting destructive electromagnetic effects, based on the transfer of the signal spectrum using wavelet transforms in special communication networks (SS SN), operating on the basis of the WIMAX standard for mobile broadband access. A Gaussian bipolar pulse is selected as a model of destructive electromagnetic influence that affects the operation of a special purpose communication network. When this type of interference is affected by a WIMAX system signal, the normal operation of the CC SN is disrupted. The MHAT wavelet was chosen as the modulating function, since it is described in the time-frequency half-plane and its parameters depend on certain coefficients of the scaling factor and the time shift. Thus, it is possible to change the parameters of the modulated signal by changing the coefficients of the wavelet function by which it is modulated. MHAT wavelet, obtained as a result of double differentiation of the Gauss function. This method finds its application primarily due to the fact that the wavelet operates in the time-frequency half-plane and its parameters depend on certain coefficients (scaling factor and time shift). Thus, it is possible to change the parameters of the modulated signal by changing the coefficients of the wavelet function by which it is modulated. The results of this method are the transfer of the spectrum of the WIMAX signal to another frequency band, where the SS SN also functions by modulating the wavelet function, as well as increasing its power and expanding the spectrum.

Keywords:broadband communication system, WIMAX communication systems, modeling of functioning under conditions of destructive electromagnetic influences, а method of counteracting destructive electromagnetic influences using wavelet transform, MHAT-wavelet.

Full text:
GilevSoavtors_2_20_1.pdf