This publication continues a series of review papers devoted to the urgent problems of designing unmanned aircrafts of the next generation, which are expected to appear in the period from 2025 to 2035. The paper considers simple and effective numerical methods based on the ideas of a vortex lift surface that allow for performing a full parametric analysis of aircrafts with very complex aerodynamic shape of the wing. Calculation examples are given for triangular wing of small and moderate aspect ratio with an acute and rounded leading edge, as well as a wing of complex shape. We consider design methods for deformed aerodynamic surfaces with optimal properties and a method for wing mechanization calculating. The formation of vortices on sharp front edges, the destruction of vortices over the wing surface and diffusion mechanism of vortices are studied. The calculation method for small aspect ratio wings is described. The results obtained are compared with the results of the experiment and numerical verification.

We have developed a method for localized laser vaporization of a sensitive element film for a solid microgyroscope based on surface acoustic waves. A double-conversion scheme was chosen as a microgyroscope sensitive element configuration. It was shown that at least three technological operations are excluded at laser vaporization as compared with photolithography method. Research results on the formation of topologies with the usage of laser processing are presented. We have determined the parameters of laser processing for sensor topology configuring by means of thin-film coating evaporation with a thickness of 400 nm, which comprises 350 nm of aluminum and 50 nm of vanadium adhesive coating applied on a substrate made of quartz. The amplitude-frequency characteristic of the manufactured sample is obtained. It is revealed that the experimental sample has a high loss value caused by inaccuracies in its manufacturing technology. Recommendations were worked out for minimization of error sources in the manufacturing technology of microgyroscope sensitive element on surface acoustic waves with the use of the laser configuration method.

Subject of Research.We study the optoelectronic system that measures vibration parameters of dither system in a ring laser gyro and serves for compensation of these oscillations. The paper considers the process of laser radiation passing through the system elements and the optical field amplitude distribution on its pattern plate. The amplitude distribution is analyzed for a small linear shift and a small tilt of the lens. Methods. We used the elements of the field calculation methodology in optical systems for information reading and writing from digital disks. Primary laser radiation parameters are determined from passport characteristics of the radiation source. The radiation transformation made by the system optical elements was evaluated by ray tracing using ray packets. The optical field amplitude distribution on pattern plate was calculated by diffraction integral with aberrations. Main Results. It is shown that the amplitude distribution is typical for diffraction limited optical system with aberrations. Simulation results give the possibility to discover that the dependences of this distribution on the shift and tilt of the lens are linear. Practical Relevance. The optoelectronic system can measure vibration parameters of dither system in real time with small errors and enables the increase of ring laser gyro accuracy.Analysis and calculation of the optical field amplitude distribution on its pattern plate is one of developing stages of the mathematical model for optoelectronic system. In future this model will be applicable for performing a detailed error analysis of the system and exploring its operation.

Subject of Research.The paper considers parameters relations in active reference mark optical-electronic system of large-sized structures spatial position control in the presence of vertical temperature gradient. We describe the idea and operation principle of the dispersion method forthe vertical temperature gradient determination bycolor camera and RGB optical radiation source. The basic condition for choosing parameters of active reference mark optical-electronic system is derived from relations that define hardware realization of the system and parameters of the air tract, which are used in the dispersion method. Method. The principle of equal influence of error components on the totalerror is used. We admitted an assumption that optical radiation is propagating in the single "refractive block" and there is no fluctuations of the air tract refractive index in time and space.Main Results.The basic criterion is proposed for choosing parameters of active reference mark optical-electronic system for large-sized structuresspatial position control in thepresence of vertical temperature gradient. The efficiency of the dispersion method for minimization  of vertical temperature gradient influence is proved. The basic condition proposed gives the possibilityto estimate this efficiency and choose the hardware parameters.Practical Relevance.The results given in the paper can be used in design of active reference mark optical-electronic systems where the influence of the air tract is strong.

Subject of Research.The paper presents the study of a graphite-epoxy composite plate strain detection possibility by the fiber-optic acoustic emission sensor mounted on its surface. Method. The proposed method consisted in additional low-frequency phase-generated carrier implementation in the impulse Fabry-Perot interferometer and its amplitude evaluation. The phase-generated carrier amplitude depends on the interferometer optical path difference, therefore, its value change can be used for the studied composite strain estimation. VCSEL with a wavelength of 1550 nm was used as a light source. The phase carrier was generated by current modulation of the light source that caused center wavelength shift of the VCSEL. Main Results. The low-frequency phase-generated carrier signal amplitude dependence on the interferometer optical path difference and wavelength shift of the light source were obtained. According to simulation results the sensitivity of the proposed method is 1.6 urad×ppm, 5.3 urad×ppm and 13.3 urad×ppm at different values of the coefficient Kd 30, 100 and 250 pm, respectively. Experimental study of the proposed method and results analysis were performed. According to experimental results, the accuracy of the proposed method was about 1´10–3% that corresponds to the sensor relative stretch of 10 me, while the accuracy of the market available fiber optic systems based on fiber Bragg grating sensors equals to 4 ppm. Practical Relevance.The proposed method can be used for strain detection of the graphite-epoxy composite constructions along with its acoustic emission control by one fiber-optic sensor.

This paper reviews the method of semiconductor laser diode frequency stabilization by phase modulation. Also parameters are identified that affect the quality of stabilization and the estimation of Allan deviation is obtained. The pilot setup has been put together and it consists of: a semiconductor distributed feedback laser diode, a fiber phase modulator, an electrical signal generator, an acetylene-13 isotope cuvette, a photodetector, a lock-in amplifier and personal computer for measurement processing. Modulated laser diode radiation passed through a gas cell provides information about the position of radiation spectral line relative to the center of gas spectral line. Gas molecular spectral lines provide frequency standard with low sensitivity to external effects. When using the reference signal, one can get an error signal in a lock-in amplifier that changes the laser diode temperature and, as a result, its wavelength. Allan deviation was estimated based on measured frequency data. Long-term stability can be improved in the time range between 0.1 s and 100 s up to 1∙10-8 (Allan deviation). This method of stabilization is useful for the development of compact high reliable optical frequency standards for space applications.

The paper considers the problem of robust control of an electric generator with unknown parameters under the effect of external bounded disturbances and the presence of high-frequency measurement noises. It is assumed that only the load angle is available for measurement, but not the rotor speed. The electric generator model is described by a system of third-order differential equations with algebraic coupling equations. The proposed approach based on the separation of the filtering and estimating properties of the control system is used to solve the problem. Unlike existing analogs this scheme gives the possibility to control independently the quality of filtering and the quality of output stabilization. Also it compensates parametric uncertainties and external disturbances with required accuracy that can be reduced by selecting algorithm parameters. The proposed algorithm is easy in implementation and selection of design parameters.

This paper describes the method of knowledge transfer between the ensemble of neural network acoustic models and student-network. This method is used to reduce computational costs and improve the quality of the speech recognition system. The experiments consider two variants of generation of class labels from the ensemble of models: interpolation with alignment, and the posteriori probabilities. Also, the quality of models was studied in relation with the smoothing coefficient. This coefficient was built into the output log-linear classifier of the neural network (softmax layer) and was used both in the ensemble and in the student-network. Additionally, the initial and final learning rates were analyzed. We were successful in relationship establishing between the usage of the smoothing coefficient for generation of the posteriori probabilities and the parameters of the learning rate. Finally, the application of the knowledge transfer for the automatic recognition of Russian conversational telephone speech gave the possibility to reduce the WER (Word Error Rate) by 2.49%, in comparison with the model trained on alignment from the ensemble of neural networks.

Subject of Research.The paper deals with the problems of the mutual reconstruction (transformation) of acoustic and visual components (modalities) of speech. Audio recording of voice represents the acoustic component whereas the parallel video recording of the speaker’s face comprises the visual component. Because of the different physical nature of these modalities, their mutual analysis is accompanied by numerous difficulties. Reconstruction methods can be used to overcome these difficulties. Method. The proposed approach is based on Principal Component Analysis (PCA), Multiple Linear Regression (MLR), Partial Least Squares regression (PLS regression) and K-means clustering algorithm. Moreover, attention is paid to data preprocessing. Mel-frequency cepstral coefficients (MFCCs) are used as acoustic features, and twenty key points, which represent the mouth contour, comprise visual features. Main Results. The experiments on the reconstruction of the mouth contour from the MFCCs are presented. The experiments were carried out on VidTIMIT dataset of audio-visual phrase recordings in English. Four variants of the proposed approach were tested and evaluated. They are based on PCA and PLS regression with clustering and without it. Quantitative (objective) and qualitative (subjective) assessment confirmed the efficiency of the proposed approach. The implementation based on PLS regression with preliminary clustering led to the best results. Practical Relevance. The proposed approach can be used to develop various bimodal biometric systems, voice-driven virtual “avatars”, mobile access control systems and other useful human-computer interaction solutions. Moreover, it is shown that, given the proper implementation, PCA and PLS reduce significantly the computational complexity of the reconstruction operation. In addition, the clustering step can be omitted to increase additionally the processing speed at the cost of slightly lower reconstruction quality.

The paper presents the study of video data encoding methods by intraframe prediction in N. 265/HEVC standard. A rapid decision-making scheme for intraframe predictions is proposed. According to the proposed scheme the complexity of the texture block coding is firstly calculated and the threshold value for rapid selection of the block coding size is set. That reduces the number of possible prediction modes. The existing and proposed encoding methods were tested on several video sequences by modeling in Visual C++ environment. To evaluate the methods criteria of video sequence quality (PSNR), bitrate and encoding time were used. The results showed that compared with the reference software of HEVC – HM 16.0, the proposed scheme gives the possibility to reduce the encoding time by approximately 13.6%, while the scheme suffers from a negligible average PSNR loss (dB 0.0033). The application of the proposed approach can reduce significantly the encoding time instead of the traditional encoding in the H. 265/HEVC video codec and can be recommended for the further study by the JCT-VC (Joint Collaborative Team on video Coding).

Subject of Research.The paper considers common natural language processing stages such as morphological, syntactic and semantic analysis. We describe connections between query parts, semantic model and internal database representation. Method. A method is proposed for the database semantic model application as a part of the natural language user database interface. There is a description of the semantic model data transforms to intermediate Q-representation of the user query. The next step is transforming of Q-representation to database SQL query based on substitution to predefined query template. Main Results. We give an account of the semantic model application as a part of natural language user interface. An example is provided of the transformation of both simple and nested natural language user queries to the database SQL query. The paper presents experimental research results of the proposed method of the semantic model application as a part of the natural language user database interface. Practical Relevance. Implementation of the proposed approach gives the possibility to resolve the problem of extracting information from the database based on user’s natural language query. This method also improves the process of resolving the ambiguity during natural language processing. The presence of information about internal database structure makes it possible to create more accurate database SQL query.

Subject of Research. The paper proposes the work organization schemes in the Industry 4.0 instrument making production companies of “smart factory” type equipped with cyber and physical systems. Studies on the production cyber-physiphication are being carried out aimed at the development and implementation of ideas and decisions in the domestic industry, which will help to establish “smart companies” in our country capable of functioning in the conditions of digital economy and possessing new production technologies. Method.We used the work organization methods for item design manufacturing with the cyber and physical systems and ontologies implementation in the conditions of the companies of the future on the basis of the design automation general theory. Main Results. It is shown that the effect of the production companies proposed organization principles implementation in the factories of the future format can boost the Russian Federation industry transfer to the most advanced digital, intellectual production technologies, robotized systems, new materials and designing methods that corresponds to the Russian Federation State Scientific and Technical Policy in the field of instrument making and machine manufacturing. Practical Relevance. The results of this research can be applied in the development of automation design algorithms for instrument making (machine manufacturing) digital production operating in the conditions of the Russian Federation digital economy.

Subject of Research.The need for slowdown of the increasing number of vulnerabilities caused by installation of unauthorized software on computer equipment, calls for an approach development to automate the audit of data storage media. The paper proposes an approach for identification of informative assembler commands. We study the effect of a chosen feature used for creation of unified program signature on the identification results. Methods. The Shannon method was used for informativity calculation. It gives the possibility to determine the feature informativity for random number of object classes and is independent of the volume of observed feature samples. Identification of elf-files was based on application of chi-square statistical homogeneity criterion. Main Results. Quantitative informativity characteristics for 118 assembler commands are obtained. The analysis of experiment results for executable files identification is carried out with the use of ten different features for creation of program signatures. Comparison is performed by chi-square homogeneity criterion at significance levels p = 0.05 and p = 0.01. Practical Relevance. We have found out the importance of particular feature application in the task of program signatures creation, as well as the possibility of considering several executable file signatures in common to create the final score of belonging to a certain program.

Subject of Research.The paper deals with the problem of development and support of educational processes through the creation of specialized semantic portals based on Linked Open Data. These portals contain the individual educational paths, and data and knowledge associated with these paths. The paper studies the problem of maintenance and support of individual educational paths in the technical domain.This isexactly the domain with the highestneed for the proposed portals. Method. Ontological engineering is chosen as a method of storage and presentation of data and knowledge for semantic portal. The basis for an educational individual path building is four-level hierarchical model, including the levels of encapsulation and composition of content, defined in terms of competences, as well as the levels of generation and composition of specific training programs. Competence modeling is performed in terms of graph theory, and SPARQL query language is used for data access. Main Results. Two semantic portals of scientific and technical information were developed with the use of the proposed methods: the first portal supports the educational process at ITMO University; the second portal supports Speech Technologies Center (STC). The results of experimental operation of the developed portals are considered. Practical Relevance. Implementing of the proposed approach and the use of the presented technique gives the possibility to develop domain semantic portals of scientific and technical information and apply them for acquiring of knowledge on individual educational paths by the user. The applied stack of technologies and methods makes it easy to adapt and reuse already developed elements of semantic portals for different domains.

Subject of Research. The paper presents the study on the JPEG image integrity violations and existing methods of their detection. We propose a method for detection of modified image and the source of its modification. The method gives the possibility to determine the original image and camera model that recorded it. Method. The method was developed with the use of machine learning tools. The following machine learning methods have been studied: naive Bayesian classifier, decision tree, logistic regression, k-nearest neighbors, SVC, random forest. The base for model training was formed by the original photos from website www.steves-digicams.comthat were modified by different graphic editors. The proposed method uses JPEG-image structure in byte view, namely, markers. Availability of markers and their number were suggested as classification features. Main Results. The trained model has demonstrated high classification result equal to more than 95%. Among all evaluated algorithms the two ones have shown the best results: decision tree and random forest. Decision tree was chosen as the best one upon stability criterion. Practical Relevance. Thereceived result can be practically applicable in the area of forensics and information security.

The necessity of service-oriented architecture application in an integrated information system design is substantiated. It is proposed to use a multi-layer data bus for the interaction of independent applications. An object-process data model is considered that implements information support for process control. A distinctive feature of the proposed model from the relational one is that the data are stored not in the form of tables, but are organized in the form of hierarchical and network structures. The data source is represented by the indicators of business processes. Business processes determine the order and procedures for the use of applications of the integrated information system. A formal mathematical description is given for the basic models of objects and processes on the basis of the predicate calculus in the form of an algebraic system. The structure of a specialized language is presented for describing and manipulating of data in an integrated information system with a Service-Based Data Manipulation Language (SBDML). The SBDML language has two components: a language for data structure describing, and a language for data retrieving and updating. The first language component contains a description of simple and complex data types. Complex types are represented in the form of tree structures. The second language component contains a description of data manipulation functions that support the interaction of applications with the data bus.

Subject of Research. The paper deals with accounts in social network websites as a source of information about the intensity of communication between employees in the team. On their basis we form success probability estimates for the spread of malefactorsocio-engineering attack on the user. Scope of Research. The research goal is to build a success assessment for malefactormulti-pass socio-engineering attack on the user based on information obtained from the accounts of company employees in social network websites which characterizes communication intensity between them. The research is aimed at development of models and algorithms for socio-engineering attack spreading on the collapsed social graph of the company and description of methods for calculation of security estimates for the information system users from multi-pass socio-engineering attacks, such attacks, where the target and the entry point do not match. Method. The methods are used of information searching, comparing and analyzing, which characterizes communication intensity between company employees, and data extracted from their accounts in social network websites. Success probability estimate of multi-pass socio-engineering attack reduces to probability estimate creation of a complex event. Main Results. A formula is presented for calculating of probability estimates of socio-engineering attack propagation between users. The estimates obtained in this way are compared to the arcs in the company's social graph, which is used in turn to assess the success probability of a multi-pass socio-engineering attack, the attack, passing through a chain of users. In the earlier studies, estimates of probabilities were defined expertly. The advantages of calculation automating of probability estimates based on data received from social network websites are described. Research Novelty.The paper considers approaches to probabilistic estimates of multi-pass socio-engineering attack success where attacks are intermediate, non-direct, and non-reducible to a single malefactoract. These estimates take into account user’s links in his or her social graph; the parameters of those links are based on the data obtained from social media/networks. Practical Relevance.The approach proposed in this paper provides the basis for further analysis of possible propagation trajectories of multi-pass social engineering attacks, as well as calculation of the probability of each such trajectory that in turn helps to expand the number of factors affecting the security evaluation of the information system users, and gives the possibility to set the backtracking task for attacks in one of the successful forms for finding solutions.

Subject of Research. The paper presents an approach that provides a numerical calculation of the absolute uncertainty of the measured physical quantity, which is capable of quantitatively estimating the minimum achievable discrepancy between the developed model and the object under study. The question of achieving a reasonable limit of measurement accuracy in engineering and science remains open, despite the use of powerful computers that take into account a huge number of quantities, and the latest mathematical methods of calculation. Since any physical and mathematical model contains a certain amount of information about the object under research, depending on the quantitative and qualitative set of the selected physical quantities, the optimal number of the chosen quantities must be found. Method. The principles of information theory are used to provide theoretical explanation and justification for the experimental results that determine the accuracy of various fundamental constants. We present the method of choosing a model with the optimal number of quantities considered and calculating the minimum achievable absolute and comparative uncertainties of the measured variable. The amount of information contained in the model is proposed as the criterion of optimality. Main Results. Within the framework of the information approach, the results show that the perfection of devices in physics and engineering is fundamentally limited. The limit of measurement accuracy calculated by the information method is much more stringent than is predicted by the Heisenberg uncertainty relation. The proposed metric provides numerical calculation of the absolute uncertainty, which is capable of quantitatively estimating the difference between the developed model and the object under study. The basic principles of the theory of measurements remain in force, and they can be used separately in the further stage of the specification and computerization of the model. Practical Relevance. The application of comparative uncertainty concept reduces the risk of selecting overestimated indicators of the equipment being designed, gives the possibility to reduce the time and cost of its development. Methodology application for calculating the relative uncertainty according to the information method will reduce significantly the financial costs of improving the International system of SI units.        

Subject of Research.Parasitic parameters, which appear on layout design stage of analog schematic model, cause negative effect on analog block performance. The presence of negative effect of parasitic parameters can be the reason for block inadequacy to its technical requirements. It leads to new design of schematic model. Since there is no automatic approach for generation of schematic models, which takes into account all types of parasitic parameters, time-consuming design efforts increase. The paper presents features of automated moving  from analog behavioral models to schematic ones. It is shown that detailed analysis of different types of parasitic parameters should be performed on schematic level to eliminate them from layout. Method. A method of analog model synthesis for digital-to-analog converter (DAC) is proposed. The proposed method improves block performance by minimization of parasitic parameters and provides technology migration. Main Results. The method contains additional stages as compared to traditional design flow: generation of schematic models from behavioral models, analysis of different types of parasitic parameters and the stage of model refinement. For implementation of the generation stage the software was designed, which performs automated generation of schematic model from behavioral description. The method was used to design 12-bit DAC on 350 nm technology. Parasitic parameters were defined and eliminated to increase the block high-speed performance. Practical Relevance. The proposed method can be used for DAC design on any technology. The method allows decreasing the effect of parasitic parameters and reducing design effort. The method gives the possibility for generating several architecture variants at once. 

Subject of Research.Timing analysis is an important stage in the design of integrated circuits. It makes possible to detect various types of errors, related both to the structure of blocks and to the violation of temporal characteristics at all levels of abstraction. It is especially important to detect errors in the interface blocks of the integrated circuit, since, otherwise, the final device may not meet the requirements of dynamic characteristics. Method. A technique for timing analysis of the integrated circuit is proposed considering the models of the contact pads and package bond wires, designed for taking into account the effect of the parameters of contact pads, bond wires, and external analog components. The technique is an extension of the standard IC verification flow with the addition of stages for analyzing and modeling of the contact pad parameters and bond wires. Main Results. The technique was used to verify the I2C interface block according to 350 nm technology. The I2C interface was included in the integrated circuit of a micromechanical accelerometer. The technique gives the possibility for detecting of errors that were not revealed during the integrated circuit design in accordance with the traditional design flow. The dynamic characteristics of the I2C controller were determined when connecting the SCK and SDA buses to the supply voltage through a resistor of 10 kΩ. Practical Relevance. The proposed technique can be used in the design of integrated circuit interface blocks. The technique takes into account the effect of the parameters of the contact pads and package of the integrated circuit, bond wires and external components on the dynamic characteristics of the circuit. It is advisable to use the technique for critical blocks, in particular, interface ones, the inoperability of which can lead to the impossibility of exchanging IC data with the consumer and, as a result, high financial costs for a new stage of IC design.

Nowadays there are a lot of continuous studies on the correct viseme classes to be used for the most effective automatic lip-reading. The paper proposes a structured approach for the development of speaker-dependent classes of visemes. This method gives the possibility to create a set of phoneme-viseme correspondence maps, where each class has a different number of visemes from two to forty-eight with a constant number of phonemes. Viseme classes are based on their mapping from phonemes, which are converted into viseme groups during speech recognition process. With the usage of the obtained correspondence maps together with the database of audio-visual Russian speech HAVRUS the paper demonstrates the dependence of the visual speech recognition accuracy on the number of used viseme classes. The application of high-speed video data made it possible to expand the optimal set of viseme classes to twenty that resulted in recognition accuracy improvement by 1.34% compared to the standard set of fourteen classes.

A method of feature combination for the problem of neural network acoustic models training is proposed aimed at the quality improvement of speech recognition. Unlike the feeding of a concatenated vector of features of a different nature to the neural network input, the proposed method uses a delayed union at the level of hidden layers. It uses individual input streams for each type of features. Such streams are able to extract specific patterns for each type of features and then combine them on the hidden layer of the neural network acoustic model. The effect of the method on the system quality was studied in the task of Russian conversational telephone speech recognition. The proposed method achieves 0.41% absolute reduction of the word error rate relative to the concatenation of features and 1.35% in comparison with the best system using one type of features. The results of the work can be used to develop automatic speech recognition systems.