Sets of vectors of decimation indices IS(id1, id2, ..., idn) of ternary M-sequences are presented, on the basis of which small and large sets of Kasami-similar sequences (KSS) with periods N = 3S–1< 20000 are formed in finite fields GF(3S) (S = 4, 6, 8). It is shown that for even values of S, the periodic cross-correlation function of a small set of KSS is three-level with the maximum value of the module of the mutual correlation function Rmax = (3S/2+1)). The correlation function of a large set at S=4 is eight-level with Rmax = (2·3S/2+1), and at S = 6, 8 is ten-level with Rmax = (3S/2+1+1). The values of the volumes of small and large sets of ternary KSS are given.

The issue of recognizing emotions in speech transcriptions, which is relevant in various fields, is studied. The influence of preprocessing methods (stop word removal, lemmatization, stemming) on the accuracy of emotion recognition in text data in Russian and English is analyzed. To conduct experimental studies, orthographic transcriptions of dialogues from the multimodal corpora RAMAS and CMU-MOSEI in Russian and English, respectively, are used. These corpora are annotated for the following emotions: joy, surprise, fear, anger, sadness, disgust and neutral. Preprocessing of text data includes removal of punctuation marks and stop words, tokenization, lemmatization and stemming. Vectorization of the resulting material is carried out using the TF-IDF, BoW, Word2Vec methods. The used classifiers are support vector machines and logistic regression. An approach is developed that is a combination of the above methods. For the Russian language, the highest accuracy of emotion recognition achieved using a weighted F-measure is 92.63 %, for the English language – 47.21 %. In addition, studies are conducted to identify the number of remote stops for effective emotion recognition from text data. Experimental results show that storing stop words in the source text allows to achieve the highest accuracy of text classification.

The problem of synthesizing an adaptive observer of state variables of nonlinear dynamic systems is considered. Correct estimation of state vector components under parametric uncertainty is a rather complex process necessary e.g. for solving several problems of systems control and diagnostic. Synthesis of the proposed adaptive observer consists of two steps. In the first one, a parameterization of the nonlinear dynamical system, which can be transformed to a state affine form, is performed. In the second step, the unknown parameters are estimated based on the gradient descent method, and a gradient-based observer for the state variables is designed.

A complete model of quadcopter motion is obtained for the problem of dynamic positioning at a point, on the basis of which two control algorithms are proposed. The first generalizes the previously obtained results to the case of a changing yaw angle. The second control algorithm solves the problem on the base of a simplified technique for tuning the controller.

A technique for measuring the birefringence of optical fibers using a Sagnac interferometer and a radiation source with a tunable wavelength is presented. Physical principles that form the basis the presented methodology are considered. To eliminate the influence of the nonlinearity of the radiation source wavelength tuning rate on the measurement accuracy, an auxiliary interferometer with known parameters is used. The obtained values of measured fiber birefringence coincide with the passport values, which confirms the proposed measurement technique effectiveness.

Results of acoustic emission monitoring of defect formation in products during direct laser growth are presented. The features of applying the acoustic emission method and results of processing recorded acoustic emission signals with the use of cascade polynomial digital filtering are considered. Results of experimental testing of the cascade filtration method for detecting internal structure defects such as cracks and pores are presented. Fragments of amplitude-time and frequency-time diagrams of acoustic emission signals recorded during the development of defects in the process of growing products are isolated. An assessment is made of the dependence of acoustic emission signals on defect formation parameters. A relationship between the acoustic emission signals parameters and applied laser radiation power is established, which characterizes the process of defect formation, as well as the nitrogen content in the heat-resistant alloy powder.

An easy-to-use method for identifying liquids and determining the concentration of substances dissolved in water is proposed, the method implementation uses short optical fiber (1–1.5 m) and does not require complex and expensive equipment. According to the proposed method, optical radiation of constant power is directed into the optical fiber, and on the same side of the fiber the power of the radiation reflected from the remote end of the fiber immersed in the liquid is recorded. It is deduced that to implement the proposed method, both single-mode and multimode optical fiber can be used using optical wavelengths of 650, 850, 1310, 1490, 1550 and 1625 nm, while the use of wavelengths 650 and 850 nm is less desirable due to greater measurement error than for other wavelengths. It is shown that the proposed method can be applied, in particular, to determine the concentration of aqueous solutions of sugar and isopropyl alcohol.

A method based on IR spectroscopy for quality control of dielectric liquids (insulating mineral oil, ester dielectric liquid Midel 7131, as well as oil mixtures with an ether content of 10, 20 and 30% vol.) during their aging, are developed and tested. IR spectra are recorded on an FT-801 IR-Fourier spectrometer manufactured by Simeks Research and Production Company. Analysis of oil samples and ester-based mixtures is carried out without prior preparation on the MFTIR (multiple frustrated total internal reflection) attachment with a zinc selenide ZnSe element and a built-in visualization system on an external monitor. Results of the research show that as synthetic ester is added to the oil, the IR spectrum of the mixture demonstrates an increase in the intensity of absorption bands caused by stretching vibrations of the bonds CO–C, C–O, C–O–C, C(=O)–O–C. The intensity of the peak characteristic of esters at a frequency of 1161 cm–1 (C(=O)–O–C) changes especially noticeably. It is revealed that during the essential oil mixtures oxidation, the intensity of characteristic bands of the carbonyl group C=O (1747 cm–1) and the ester group C(=O)–O–C (1161 cm–1) decreases, while the absorption intensity in the frequency range of the CH group (2920 cm–1) remains practically unchanged. It is proposed to evaluate the degree of decomposition of mixtures of ester dielectric liquid, as well as essential oil mixtures, using the “destruction index”, which is calculated using a formula that takes into account the change in optical density in the band of the C=O and C(=O)–O–C groups. This indicator is anticipated to allow for quantitative control of the quality of alternative ester liquids and essential oil mixtures during their service.

A study is conducted to obtain spectral data on the penetration of the chemically cured adhesive “Tokuyama Universal Bond II” into dentin using various techniques. An external view of the used spectrophotometer is shown, an optical diagram of the instrument is presented, and the technical characteristics of the device are formulated. Samples for research, prepared using various methods, are depicted. The spectral dependences of the absorption coefficient are obtained in the zone of the boundaries of fixation of composite inlays with dentin and the border areas of the specimen-sections when using the Tokuyama Universal Bond II adhesive according to the protocol described in the manufacturer's instructions: when applying “Tokuyama Universal Bond II” with an additional 10-second exposure on the dentin before swelling begins; by rubbing “Tokuyama Universal Bond II” into the dentin surface for 10 seconds.