An approach to the construction of a code with the summation of the bits of information vectors in a ternary asymmetric number system is proposed. Such codes are focused on the detection of errors in information vectors and have a relatively low redundancy, which makes it possible to use them effectively in the development of traceable digital devices and systems with fault detection. The features of error detection by the presented code with summation are analyzed in comparison with the known code with summation, which implies separate control of digits equal to one and digits equal to two. It is shown that a code with direct summation of the bits of information vectors uses its control bits more efficiently than the code considered earlier; moreover, it has less redundancy. At the same time, however, the proposed code does not reveal much more errors in information vectors, which can be both compositional and monotonic and asymmetric. The above circumstances are recommended to be taken into account when synthesizing digital devices operating in ternary logic.

A mathematical model of an enterprise information-computing network functioning with the choice of the optimal policy for the work of the information technology department is presented. The task of the research is to find the optimal strategy for the work of the information technology department to improve the information and computer network performance. The information-computer network of an enterprise is considered as an example of Markov network of mass service. The choice of the best strategy to maximize the mathematical expectation of a payoff is mathematically substantiated. The methods of probability theory, the mathematical apparatus of the theory of mass service and the theory of controlled Markov random processes are used as a theoretical apparatus for studying the system under consideration. Howard's iterative algorithm is used for finding the optimal policy.

The issues of assessing the functional state of network infrastructure objects are considered, and the need for using alternative assessment methods is justified. Side channels are described, with the use of which it becomes possible to monitor the states of individual objects. The experiment conducted in the framework of the study is described. With the help of two sensors, statistical information about the various operating modes of the network architecture device was read. The obtained data were processed using two-layer direct propagation neural networks with a sigmoidal transfer function in hidden layers. The resulting behavior model of the network infrastructure object can be used as an additional element for determining the functional state.

A method for estimating parameters, which, in turn, can be represented as outputs of linear generators. In general, these are non-stationary parameters described by polynomials of time, as well as sinusoidal and exponential functions of time with unknown amplitudes and phases. In this paper, attention is paid to the case of polynomial parameters. The solution to the problem is based on transforming the object model to the form of a linear regression equation with respect to the state variables of the generators, the outputs of which describe the sought parameters. The method of dynamic expansion and decomposition of the regressor (or mixing of the regressor) makes it possible to solve the problem of restoring all state variables and output signals of the mentioned generators in a finite time. A numerical example of identifying parameters of a model of surface vessel motion along the course is presented.

Several technological approaches to reduction of surface defects in polymer planar optical waveguides are presented. Experimental studies are shown to reveal and eliminate the factors contributing to an increase in the number of surface defects when using soft lithography technology. The efficiency of using a specially designed light filter is demonstrated.

A device for cooling elements of radio-electronic equipment operating in the mode of intermittent thermal loads is considered. The device uses melting working substances with an additional air heat sink due to fan units. An experimental stand is described, results of the device field tests are presented. Time dependences of the container shell temperature during melting and solidification of the working substance, as well as duration of complete melting (solidification) of the working agent caused by supplied (removed) heat, are presented. It is determined that the use of fan assemblies makes it possible to reduce the substance solidification time by about 1.7 times in relation to the case of natural heat exchange with the environment by means of a copper radiator. According to estimates carried out, the discrepancy between the experimental and calculated data is 8 %.

Over the past decade, the quality of natural language processing has improved significantly due to enhancement in both software and hardware. Modern technologies demonstrate high results in all areas - from automatic speech recognition to text-to-speech and natural language understanding. This allowed a large number of companies to significantly increase the speed of service delivery, quality and usability of products in general. However, there are many other not yet used applications of voice interfaces, and one of them is the application of such technologies in industrial production within the framework of modern enterprises. There are a number of challenges that need to be addressed when using voice technology. The basic components of natural language processing are presented, the features of their application, as well as approaches to creating an effective and at the same time simple voice user interface. Some features of industrial design and production are considered, which should be taken into account when developing such systems.

The object of research is profilograms (random vectors) of two functional metal surfaces that were in frictional interaction. Possibilities of analysis of the main components using the properties of covariance matrices are determined. The evaluation by the sample covariance matrix is carried out. The eigenvectors and eigenvalues of the covariance matrix are calculated. The coordinate system defined by the eigenvectors makes it possible to estimate the maximum and minimum variance. The evolution of microgeometry variability is traced at different times of interaction of tribopairs. Thus, a DGC analysis technique has been proposed.

Unmanned aircraft systems make it possible to perform routine and emergency diagnostics of electric power facilities faster and cheaper than traditional methods of inspection with the help of teams of workers. Transition to the new technology requires a deep study of unmanned aerial vehicles available on the market and selection of optimal devices for diagnosing various elements of power grids. There are many ways to obtain information on defects in electrical power facilities. The inspection results largely depend on the equipment used in the unmanned aircraft system. Various types of devices designed to obtain information from unmanned aerial vehicles are considered, and conclusions are drawn about their optimal choice for solving certain problems of diagnosing of power grids elements.

The possibility of evaluating the quality indicators of cylindrical gearing by the free run-out method, which consists in measuring the rotation time of a gear train under the action of an inertial load, is considered. A design of a universal stand for the implementation of this method is proposed and its structural diagram is presented. The principle of the stand operation is briefly described, and its main advantages are given.