Methods for studying time characteristics of corporate computer network functioning are considered. Queuing models are used that allow to account for the random nature of the processes occurring in the network. The influence of priorities between packages of different classes on the quality of their transmission is shown.

The Smith scheme of including a consecutive compensator in the direct branch of a designed system is considered. The scheme is used as an effective way of solving the problem of removing the link of pure delay beyond the closed loop of the system to ensure its stability regardless to the delay magnitude. The system property is based on additive signal compensation, which is not a robust procedure. A correct implementation of the procedure requires an accurate knowledge of the delay value and ordinary parameters of the polynomial part of the plant transfer function. Sensitivity of processes in the projected system to the deviation of ordinary parameters from their nominal values is analyzed.

A new approach to synthesis of modal control of linear system with single input and single output (SISO) is proposed. Stabilization of an unstable object is realized with formalized matrices by means of modification of interval factors of initial matrix in the limited range of their variation. The developed technique allows to carry out displacement of roots of the characteristic equation corresponding to poles of a transfer function, from the right to the left half of complex plane thereby modifying the system dynamics and converting it from the unstable to a steady condition. The difference of matrices corresponding to the new steady and initial unstable system can be identified with the concept of modal control. From the steady formalized matrices received in a piecewise continuous cycle of their factors modification, variety of pointwise and interval matrices of modal control is drawn up.

The problem of determining the exact characteristics of demands entering as a lot (group) into system with radial structure is investigated. The consideration accounts for the inhomogeneity of processing modules and temporal transitions determined by the specifics of the common resource (transport robot) activity. The problem is solved by drawing up a system of time event equations that reflect the transition times of demands from one module to another, in the class of backward sequence dispatching discipline. A method for calculating the characteristics of the cluster is developed. The characteristics include the time of the first delay, the cycle of demands, the number of demands in the lot group with the account for reconfiguration of the cluster and the planned restrictions on the time of the lot.

A new method is proposed for organization of concurrent error detection system of combinational logical circuits with 100 % detection of single stuck-at faults on inner structure elements outputs. The method is based on partition of circuit outputs into the groups in which values of maximum two outputs may be distorted simultaneously. Module-weighted Berger codes with data vector lengths m = 2b, b = {2, 3, …} are used for the control. It is noted that module-weighted code with m = 4 is the most effective for this purpose: its generator has the simplest structure for which the self-checking property is easily ensured.

Based on developed algorithm for generating Gordon — Mills — Welch sequences, a full list of testing polynomials for GMW-sequences with the period N = 511 is obtained. Binary GMW-sequences are formed on the basis of base M-sequence over finite fields with double expansion GF[(2m)n] and can be represented as a matrix of dimension [JxL]=[(2m–1)x(2m+1)]. A qualitative specifics of sequences with the period N = 511 consists in the fact that they are formed over a finite field GF[(23)3] and are presented in the form of a matrix of dimension [JxL]=[7x73], but not in the form of quasi-quadratic matrix of dimension [(2m–1)x(2m+1)]. Equivalent linear complexity of these sequences corresponds to the degree of the testing polynomial which can be represented as a product of three irreducible polynomials of the ninth degree. GMW-sequences with period N = 511 are formed using M-sequences of the same period. There are 48 primitive polynomials of the ninth degree in the field GF(29), the full list also contains 48 test polynomials for GMW-sequences. 

An algorithm of spectral analysis of measuring signals allowing for signal processing within the interval of observation is proposed. Effectiveness of the algorithm application is confirmed by presented estimates of the error in numerical determination of the Fourier coefficients. With the use of Bernoulli polynomials, the exact upper bound of error of the algorithm is evaluated. The obtained generalized expression for the error depending on the differential and spectral properties of the signal, as well as the acceptable error values allow for determining the parameters of the signal sampling.

Machine vision systems are a dynamically developing field of robotics. They give the ability to detect, visualize, track and recognize objects to the manufacturing and controlling cyberphysical systems. The use of such systems along with modern image processing algorithms allows shifting part of the operator's routine duties to a robotic system, in accordance with the industry 4.0 paradigm. An important property of the machine vision system is the resolving power, which can be estimated using various parameters and characteristics. The goal of this paper is to compare existing resolution measurement methods for the machine vision systems and to discuss their advantages and drawbacks.

Analysis of Michelson interferometer operation in two measurement modes: touch-trigger and scanning, is carried out. A method is proposed for eliminating measurement errors introduced by longitudinal component of the speckle structure when the instrument is in trigger mode. The case of a relatively slow transverse component of random modulation of object signal is considered; the component is responsible for the emergence of the longitudinal component leadings to increased measurement errors. Requirements to the elements and settings of micro-scan are formulated.

Main stages of modeling a high-power DC-DC converter are considered. The investigated transducer has supply voltage varying in the range 175-320 V, and the output voltage of 610 V; it consists of a bridge inverter raising a high-frequency transformer and a diode rectifier loaded on a capacitive filter and an active load. The presence of the transformer leakage inductance makes it possible to provide soft commutation of power switches. A nonlinear model of the converter is constructed with the use of the SimLower Simulink extension package of the MatLab system, main stages of the model synthesis are presented. The main problems that can arise in the simulation and transition from the theoretical study of a high-power converter to experimental design are indicated. The main parameters of the nonlinear model are chosen, and their effect on the energy loss is analyzed. A comparative analysis of the characteristics obtained in the simulation and the characteristics of the real device is carried out.

Computer emulations of visual simulators are considered. Organization on a personal computer of cyclic slide shows that exert a simulator effect on the tone of ciliary intraocular and external oculomotor muscles, is described. The impact on the visual cells (retina) by the color impulses of a healthy person when viewing a slide show is quite like the effect produced on the retina by devices for correcting the human body functional systems, developed under the guidance of T. P. Teterina, Academician of the RAS.