Methodological foundations for creation and use of integrated system of decision-making support aimed at formation and presentation to the decision maker of ordered options for control actions on certain complex objects functioning in various areas are presented. Various methodical approaches to synthesis of integrated system of decision-making support are analyzed. The developed methodology for solving the problem under study is based on results obtained in an interdisciplinary field of system knowledge, within the framework of such scientific areas as systemology, neocybernetics, and informatics. It is proposed to use three fundamental systemic-cybernetic concepts as basic concepts for the construction and use of integrated system of decision-making support: the concept of complex (systemic) modeling of complex objects, the concept of proactive control of their structural dynamics in changing conditions caused by the influence of the disturbing environment, as well as the concept of intellectualization of control, which provides for the need to use new intelligent information technologies aimed at achieving complex integration of natural and artificial intelligence. Basic principles of solving the formulated problems and the structure of the choice of the solution in the conditions of uncertainty and multi-criteria are proposed.

An approach to development of software and mathematics that allows to solve the problem of synthesizing an integrated plan for modernization of an information system and for a complex object functioning is developed. A model of production processes and a management model of information services modernization are presented. A software prototype is described that synthesizes the optimal integrated plan based on the Boltyansky local section method. The developed software package makes it possible to rationally approach the choice of the initial dispatch plan and find compromise solutions in the presence of many possible options for modernizing integrated information systems. Results of experiments are presented.

A new fuzzy-probabilistic approach is proposed for solving the problem of monitoring a complex object state. The approach provides the possibility to extract, represent, formalize, and employ explicit and implicit expert knowledge about such objects using analytical expressions. The proposed approach combines elements of the theory of fuzzy sets in terms of expert knowledge representation and formali-zation by the methods of the theory of experiments planning when assessing the state of complex object. A complex object functioning is characterized by the presence of both measurable and non-measurable (organoleptic) information, which significantly complicates creation of a mathematical apparatus for con-trol systems. All stages of solving the problem of constructing a fuzzy-probability model are considered, an example of using the proposed approach for assessing of a specific complex object state is given.

A method for monitoring the state of onboard systems of spacecraft, based on the basic principles of a synergistic approach to spacecraft flight control, is considered. For synergistic observations of the onboard systems state, the postulate is introduced about the interconnection of all processes occurring on board. On its basis, a search is carried out on board the spacecraft for such physical processes in its main onboard systems, the measured parameters of which allow, taking into account external factors, to assess the parameters of the technical state of adjacent systems, as well as to analyze resources consumption in the main and adjacent systems. As an example, a method is considered for determining the magnitude and localization of the intrinsic magnetic moment of solar batteries based on measurements and calculations of the spacecraft angular velocity and the kinetic moment of power gyroscopes in the motion control system with the account for external disturbing moments.

The problem of analysis of telemetry data of small spacecraft onboard equipment aimed at identification of its functioning state, is considered. Algorithms of stream clustering in solving this problem are studied. It is noted that the use of such algorithms makes it possible to single out the cluster data structure, as well as to trace its dynamics together with the automatic detection of abrupt changes associated both with a change in the state of onboard equipment systems functioning, and with the possible appearance of failures in their operation.

A unified virtual electronic passport of rocket and space technique products is presented. The substantive part of the electronic passport logically consists of two parts - the main one, which includes information that provides accounting, control and analysis of the technical condition of the product, and the auxiliary one, containing reference data in the composition and volume necessary to support the product during operation. It is emphasized that the unified virtual electronic passport should be relevant throughout the entire life cycle of the product due to the regular and timely entry of relevant records on operational and / or repair work into it and should be generated automatically or semi-automatically. New information technologies are developed to be used in creation and use of the electronic passport, which is an automated geographically distributed information system accounting for the product design, creation, testing and operation features, as well as collection, processing, storage and presentation of generalized information about technical state and reliability of space facilities. The specific capabilities of the created unified virtual electronic passport are illustrated by the example of the Soyuz-2 space launch vehicle.

Specifics of communication infrastructure based on a content-controlled network are discussed. The purpose is to solve the problem of constructing an information and communication environment that provides data transfer from sources to consumers and allows for the possibility of software reconfiguration. An approach to virtualization of the data transmission medium is developed, based on encapsulation of related components, which makes it possible to consider the data transmission channel as a service or a virtual entity including a transmission medium, a set of hardware and software tools for transforming data flows between different physical environments, hardware and software for controlling transport streams, as well as hardware and software for encoding-decoding content. A distinctive feature of the approach is the use of active data technology (code fragments transmitted over the network and executed on network nodes), which enables the virtual data transmission medium to be considered as a distributed virtual machine, while the virtualization process is controlled by active data. To formalize the approach, multilayer graphs are used and modified to describe not only the network topology, but also the internal structure of nodes based on computing (software or hardware) components and interfaces. Using the proposed approach, a functional diagram of the implementation of a distributed information and communication environment based on a system of virtual machines is developed. Variants of interaction between nodes and distribution of executable code among network nodes in such an environment are considered and analyzed.

The problem of improving the architecture of an integral neural-network model of Russian speech recognition is discussed. The considered model is created by combining the codec model with the attention mechanism, and the model based on the connectional temporal classification. Application of such neural network architectures as Highway Network, residual connections, dense connections, in the end-to-end model is studied. In addition, the use of the gumbel-softmax function instead of the softmax activation function during decoding is investigated. The models are trained using transfer learning method with English as non-target language, and then trained on a small corpus of continuous Russian speech with duration of 60 hours. The developed models are reported to demonstrate a higher accuracy of speech recognition in comparison with the basic end-to-end model. The results of experiments on recognition of continuous Russian speech are presented: the best result is 10.8% in terms of the number of incorrectly recognized characters and 29.1% in terms of the number of incorrectly recognized words.

A model of a legal document of the “contract” type is built and used as the basis of a system developed for legal expertise automation. The existing methods of automatic processing of texts of legal documents are analyzed, their specificity is determined. To accomplish the task, an associative-ontological approach is used, and methods of text summarization are applied. To simplify the legal examination, the text of the agreement is presented in the form of a non-strict sequence of text blocks, each of which reflects a semantic load independent of other blocks. The problem of highlighting typical sections from the text, described by means of a set of mandatory and variable blocks in the order of their placement in the contract, is considered. A system for the text blocks selection is been developed based on the methods of summarization and associative-ontological representation of sentences. An algorithm for correlating sentences or their parts to one of standard blocks is proposed. The resulting model is planned to be used for processing agreements of the "consent to the processing of personal data" type.

An approach to formation of requirements for the design process of secure cyber-physical systems is described. This approach covers one of the stages of the design and verification methodology for such systems. During this stage, the customer's wishes are transformed into specific requirements and constraints, which determines the design process. The transformation is performed based on the formed knowledge base. As an example of the approach application, the process of forming requirements for the design of a mobile robot for an object perimeter monitoring is presented.

Results of experiments on testing the perception of 12 visualization components in virtual reality based on blind testing of 20 users, are presented. The tests are carried out using HTC Vive VR glasses and controllers. An assessment of the speed and accuracy of visualization perception of 9 quantitative and 3 categorical components is performed. Based on the upper quantile of the distribution of estimates, the components recommended for use in visual analytics systems are determined. The obtained estimates of the speed and accuracy of perception allow to account for the features of perception when designing analytic systems using virtual reality.