The practice of deploying applications in the form of containers, which has become widespread in recent years due to its high efficiency and flexibility in managing computing resources is discussed. Containerization allows accelerating the development and deployment of software, increases the stability, performance and reliability of applications during adaptive reconfiguration of the system to changes in request flows and accumulation of failures. When substantiating the choice of design solutions, the efficiency of cluster systems with container virtualization is estimated by the profit received from operating the system per unit of time. This profit depends on the impact of the number of deployed containers on request servicing delays and on infrastructure support. The difficulty of estimating the average wait time for requests in the cluster nodes is due to the complex effect of dynamic sharing of shared computing resources of nodes between active containers on their performance during container virtualization. The study is aimed at improving the efficiency of cluster container virtualization systems with adaptive reconfiguration of the number of nodes and containers deployed in them depending on the intensity of requests coming into the cluster. The choice of cluster configuration is made taking into account the reduction of operating costs and increase in profit from the timely and error-free execution of functional requests.

The relations for determining the equivalent linear complexity (ELC) l_S of non-binary Gordon — Mills — Welch sequences (GMWS) formed in arbitrary extended finite fields GF[(p^m)ⁿ] are presented. The values of the ELC of the GMWS for fields with a base p = 3 -17 are obtained, taking into account the parameter М_n(r_p) equal to the number of summable sequences during the formation of the GMWS. It is shown that the parameter М_n(r_p) depends exclusively on the degree n of the field expansion and the values of the digits of the p-ary representation of the number r_p, which is mutually simple with the order of the multiplicative group of the subfield GF(p^m).

One of the possible directions of increasing the noise immunity of systems using the direct sequence method for spectrum spreading is investigated, namely, a change in the paradigm that assumes that code sequences should be binary and symmetric, in favor of non-binary and asymmetric sequences. An approach to the formation of ensembles of quasi-orthogonal code sequences with high structural secrecy is presented. The specified characteristics are achieved through the analysis of known Gordon — Mills — Welch (GMW) code sequences with good correlation properties and high structural secrecy based on the theory of quasi-orthogonal matrices. These sequences are the basis for constructing cyclic Mersenne matrices with elements {1, –b}. The prototype, the GMW sequence, is modified by replacing the element “0” with the element “–b”, which is calculated in accordance with the theory of quasi-orthogonal matrices. Autocorrelation and intercorrelation functions are calculated for the resulting ensemble. It is shown that quasi-orthogonality of the sequence ensemble being formed is achieved, and at the same time the correlation properties are not worsened in comparison with the prototype. The obtained results have both independent significance and can be a component of the algorithms for generating GMV sequences.

An adaptive observer for a bilinear nonstationary dynamic system under partial parametric uncertainty is proposed. The problem is solved under the assumption that the unknown parameters are contained in the matrix/ vector at the control signal. The key idea of the proposed algorithm is a new parameterization of the object based on two functions, one of which can be found analytically using the known and measured signals of the system. The use of linear filters allows us to reduce the system to the form of a linear static regression model containing unknown constant parameters; at the next stage, the unknown parameters are estimated using a gradient algorithm. Since the unknown constant parameters are mathematically related to the unknown initial conditions of the state vector and the unknown variable parameters in the matrix/vector of control, the estimates of the unknown components of the state vector and the estimate of the unknown parameter are derived based on the estimates obtained. It is shown that the proposed method advantage consists in the possibility of application to a sufficiently wide class of bilinear systems, to which, in particular, Euler-Lagrange systems describing many real technical objects and robotic systems can be reduced.

The problem of synthesizing a fast decentralized control algorithm for uniform distribution of a group of mobile robots making up a multi-agent system, over a given straight line segment is considered. The aim of the study is to synthesize a group finite control algorithm to achieve distributed consensus and subsequently test it experimentally. The method used is based on the homogeneity theory and the implicit Lyapunov function method. The stability criteria for a multi-agent system are specified as a solution to a system of linear matrix inequalities. Unlike the existing method, the proposed approach allows weakening the stability criteria, simplifying the setup and application of the algorithm, as well as effectively analyzing the system dynamics and guaranteeing consensus in a finite time. The theoretical results are experimentally confirmed using an IP camera and 4 wheeled robots. The experiments demonstrated high efficiency of the proposed approach, which confirms its applicability to control problems in real conditions.

The process of formation of a multilayer Al-Ni structure by the magnetron sputtering method used for the reaction of self-propagating high-temperature synthesis is considered. The magnetron sputtering system arrangement is proposed, which allows using six magnetrons and an ion source in a single technological cycle of formation of a multilayer Al-Ni structure, which leads to an increase in the rate of its growth. A sample of the obtained Al-Ni structure is studied using scanning electron microscopy. It is shown that the sample has a thickness of 50 μm, the thickness of one bilayer is 80 nm. The results of studying the obtained structures using Auger spectroscopy and scanning electron microscopy are presented, and an elemental analysis of the formed multilayer structure is carried out.

Approaches to the prediction and diagnosis of diseases of the cardiovascular system are discussed. The main research method is process analysis technology — process mining, which is most effectively implemented in the ProM framework. To analyze digital cardiogram data, the Multi-perspective Process Explorer plugin is used, which allows one to compare a reference model with an event log and provides tools for analyzing the data obtained. As a result of the research, process models reflecting deviations from the standard have been obtained, on the basis of which the user can conclude about of the cycle duration, the duration of each peak and complexes. The results of the study confirm the expediency and effectiveness of using the Multi-perspective Process Explorer plugin for the task of analyzing digital cardiac signal data.

The realization of designing in CAE application of design solutions for optoelectronic systems node and its manufacturing technology in CAM environment is discussed. Stable functioning in a wide temperature range of information optical devices for environmental control and measurement of physical parameters is achieved at the design level by measures aimed at reducing the influence of thermal deformations. At the technological stage of production this problem is solved by assigning special methods of metal processing in order to minimize residual mechanical stresses in the parts of the device assemblies. Automation of design and technological projects is carried out on the basis of CAD/ CAE/CAPP/CAM platforms.

The problem of searching for a route of movement of a robotic system in a space with obstacles is considered. As a basis for solving the problem, a class of algorithms for path planning with a random sample of analyzed points of space is selected. A modification of the bidirectional fast-following random tree algorithm with a dynamically changing search zone, which is determined geometrically, is presented. The presented modification allows finding shorter routes of movement and reducing the load on computing resources compared to previous modifications.

The development of control and sensing algorithms for collaborative mobile robots with hybrid holonomy is discussed. A mobile platform with a roller locking device providing transition between omnidirectional and all-wheel drive modes is considered. The system structural diagram, control algorithms for the motors and the locking device, and the method for processing data from sensors are presented.